The Population Status, Habitat Use and Seasonal Diet of African Elephant (Loxodonta Africana) in Majete Wildlife Reserve, Malawi

The population status, habitat use and seasonal diet of African elephant (Loxodonta africana) in Majete Wildlife Reserve, Malawi

Frances Ann Forrer

Thesis presented in partial fulfilment of the requirements for the Degree of Master of Science, Department Conservation Ecology & Entomology, Stellenbosch University

Supervisor: Dr Alison Leslie

Faculty of AgriSciences Department of Conservation Ecology & Entomology

March 2017 Stellenbosch University https://scholar.sun.ac.za

Declaration

By submitting this thesis electronically, I declare that the entirety of the work contained therein is my own, original work, that I am the sole author thereof (save to the extent explicitly otherwise stated), that reproduction and publication thereof by Stellenbosch University will not infringe any third party rights and that I have not previously in its entirety or in part submitted it for obtaining any qualification.

Frances A. Forrer

March 2017

i Stellenbosch University https://scholar.sun.ac.za

Abstract

The African elephant (Loxodonta africana) is classified as a keystone species as it is critical to the integrity of the ecosystems it occupies. It influences a variety of factors in these ecosystems that include, but are not limited to, canopy cover, seed dispersal and various plant and animal species distributions. In addition to being classified as mixed feeders, elephants are water-dependent and the location and availability of water affects the extent and intensity at which elephants make use of vegetation. Confinement through the fencing of many elephant populations, particularly in Southern Africa, has adversely affected the management of this species. Population numbers tend to rapidly increase due to improved protection and supplementation of resources, intensifying the species negative effects on other herbivore species. Majete Wildlife Reserve, located in Malawi, was almost entirely devoid of wildlife but was revived by African Parks in 2003. The reserve was fenced, artificial waterholes were installed and an abundance of wildlife was reintroduced, including 213 elephants. Five years post reintroduction, elephant numbers have dramatically increased and concern has been raised regarding the potential impact of this species on the vegetation and other herbivore species in the reserve. In this study, a review of all relevant literature was reported and two field studies were conducted on the population status, habitat use and diet of elephants in the reserve.

The population status of the elephants was assessed with aerial survey data and individual identification techniques. The population has increased to an estimated 389 individuals, of which 366 were positively identified. Results revealed a sex ratio of 1:1 and a population growth rate of 13.8% per annum. Additionally, habitat use of the elephants was investigated using camera trap gridlines throughout the reserve. It was determined that a higher number of elephant frequented habitat near perennial water sources and at lower altitudes. Furthermore, waterhole usage was determined using camera traps placed at artificial water sources in Majete. Results suggested that fewer elephants utilised artificial waterholes during the wet season and that family herds tended to dominate the use of the majority of the artificial waterholes. Lastly, it was determined that the use of artificial waterholes was increasingly homogenous in the dry season. The increasing elephant population resulted in dispersal to less preferred areas, namely that of higher altitude miombo woodland as lower altitude regions were potentially becoming too densely populated.

Diet of the elephants was investigated using stable isotope analysis of faecal samples to determine seasonal grass and browse composition. Elephants’ diets displayed a clear seasonal difference in the proportion of C3 browse consumed. In the dry season the diet contained 98% C3 browse but decreased to 59% in the early wet season and to 65% in the late wet season. This indicates that a

Stellenbosch University https://scholar.sun.ac.za

greater proportion of C4 grass was consumed in the wetter seasons, typical of other elephant populations.

The results from this study will contribute towards the compilation of an elephant management plan that will be provided to African Parks, Majete, for further implementation.

iii

Stellenbosch University https://scholar.sun.ac.za

Opsomming

Die Afrika-olifant (Loxodonta africana) word geklassifiseer as ’n hoeksteenspesie omdat dit van kritiese belang is vir die integriteit van die ekosisteme waarin dit voorkom. Dit beïnvloed ’n verskeidenheid faktore in hierdie ekosisteme, insluitend, maar nie beperk tot, lowerdekking, saadverspreiding en die verspreiding van verskeie plant- en dierspesies. Buiten dat hulle as gemengde vreters geklassifiseer word, is olifante afhanklik van water en die ligging en beskikbaarheid van water beïnvloed die mate en intensiteit waartoe olifante van plantegroei gebruik maak. Die inperking van baie olifantbevolkings deur heinings, veral in Suidelike Afrika, het ’n negatiewe effek op die bestuur van die spesie gehad. Bevolkingsgetalle neig om vinnig toe te neem as gevolg van verhoogde beskerming en die aanvulling van hulpbronne, wat die spesies se negatiewe effekte op ander herbivore versterk. Die Majete Wildreservaat, wat in Malawi geleë is, was feitlik gestroop van alle wild, maar is in 2003 deur African Parks herleef. Die reservaat is omhein, kunsmatige watergate is geïnstalleer en ’n oorvloed wild is hervestig, insluitend 213 olifante. Vyf jaar ná die hervestiging het olifantgetalle dramaties toegeneem en kommer is uitgespreek oor die potensiële impak van hierdie spesie op die plantegroei en ander herbivoorspesies in die reservaat. In hierdie studie word verslag gedoen oor die relevante literatuur, asook van twee veldstudies wat onderneem is om die bevolkingstatus, habitatgebruik en dieet van die olifante in die reservaat te ondersoek.

Die bevolkingstatus van die olifante is geassesseer deur gebruik te maak van data afkomstig van lugopnames en individuele identifikasietegnieke. Die bevolking het toegeneem tot ’n geskatte 389 individue, waarvan 366 positief geïdentifiseer is. Die resultate toon ’n geslagsverhouding van 1:1 en ’n bevolkingsgroeikoers van 13.8% per jaar. Daarbenewens is die habitatgebruik van die olifante ondersoek met behulp van kamerastrikke op ruitlyne in verskeie plekke in die reservaat. Daar is bepaal dat ’n groter getal olifante die habitat in die nabyheid van standhoudende waterbronne en op laer hoogtes bo seespieël meer dikwels besoek het. Die gebruik van watergate is verder ondersoek met behulp van kamerastrikke in die nabyheid van die kunsmatige waterbronne in Majete. Die resultate stel voor dat minder olifante die kunsmatige waterbronne tydens die nat seisoen gebruik het en dat familietroppe geneig het om die gebruik van die meerderheid kunsmatige watergate te domineer. Laastens is daar bepaal dat die gebruik van kunsmatige watergate toenemend homogeen was in die droë seisoen. Die groeiende olifantbevolking het gelei tot verspreiding na minder gunstige gebiede, veral na die hoërliggende miombo bosland soos die laer geleë streke potensieel te dig bevolk geraak het.

Die olifante se dieet is ondersoek met stabiele isotoop analise van mismonsters om die samestelling van seisoenale gras en takvoer te bepaal. Die olifante se diëte het ’n duidelike seisoenale verskil

Stellenbosch University https://scholar.sun.ac.za

getoon in die proporsie C3 takvoer wat gevreet is. In die droë seisoen het die dieet 98% C3 takvoer bevat, maar dit het afgeneem tot 59% in die vroeë nat seisoen en tot 65% in die laat nat seisoen. Dit dui daarop dat ’n groter proporsie van C4 gras in die natter seisoene gevreet is, wat tipies is van ander olifantbevolkings.

Die resultate van hierdie studie sal bydra tot die samestelling van ’n olifantbestuursplan wat aan African Parks, Majete verskaf sal word vir verdere implementering.

Stellenbosch University https://scholar.sun.ac.za

Acknowledgements

First, I would like to thank my supervisor Dr Alison Leslie for giving me the opportunity of a lifetime, for guiding me with this project and for being so patient with the thousands of questions I had. Thank you for your enthusiasm and your constructive input, you truly helped me make this masters a success. I am most grateful to Dr Frans Radloff for providing me with invaluable advice on chapter 4, isotopic analysis of elephant faeces. I would also like to acknowledge Dr Dan Nel who helped me with my statistics. Lastly I would like to express my gratitude to the Earth Watch Institute for funding the Animals in Malawi Research Project, including my own research. Thank you for sending us the most amazing volunteers, they were truly a wonderful array of people who were passionate about Majete and the work we do.

I would also like to say a great big thank you to African Parks Majete for providing me with this research opportunity and for supporting me during my fieldwork. Thank you to Craig Hay, Gervaz Tamala, Patricio Ndazela and Tizola Moyo for your advice and support. Your enthusiasm and willingness to promote research in Majete is inspiring and I truly hope this particular study is able to provide you with the necessary information you need to manage and conserve Majete’s wonderful elephants. A special mention to Mr Isaac Mulilo and the workshop team who literally kept our vehicles going, I am not sure what we would have done without you. I would also like to express my gratitude to Martin Awazi and the scouts of Majete who kept us safe on long field walks, your energy, good humour and patience was much appreciated. Lastly, thanks must also go to the Majete staff body who were so welcoming to me during my long stay.

To my parents, I could not have done this without your love and support. I am fairly sure I gave you both sleepless nights with my adventures in Malawi but thank you for having faith in me and encouraging me. To my dear sister, Heather, thank you for always being there for me and for your wonderful visit to Malawi, it is one I will never forget. To Emile, thank you for your love and support, I will always be eternally grateful that I bumped into you in the middle of Majete. To my fellow researchers: Fafa, Willem and Charli thank you for all the good times we shared. Lastly, to my dear friend and fellow Majete girl, Claire Gordon, you were my rock and we made it through the thick and thin together, so thank you so much for the adventure!

Stellenbosch University https://scholar.sun.ac.za

Table of Contents

Declaration ...... i Abstract ...... ii Opsomming ...... iv Acknowledgements ...... vi Table of Contents ...... vii List of Tables ...... x List of Figures ...... xi List of Abbreviations ...... xii

Chapter One: Thesis Introduction and Outline ...... 1 Introduction ...... 1 Research Question ...... 5 Research Statement ...... 5 Hypotheses ...... 6 Alternative Hypothesis ...... 6 Null Hypothesis ...... 6 Research Goals ...... 6 Research Objectives and Research Questions ...... 7 Significance of Research ...... 7 Scope of Limitations ...... 9 Assumptions ...... 10 Brief Chapter Overview ...... 11 References ...... 12

Chapter Two: African elephant (Loxodonta africana) population demographics and resource use in protected areas: A review ...... 16 Abstract ...... 16 Introduction ...... 17 Elephant Demographic Vital Rates and Population Growth ...... 19 Elephant resource use and range distribution in protected areas ...... 23 Geographic Variability ...... 26 Seasonal Rainfall and Habitat Selection ...... 27 Sexual and spatial segregation in habitat use ...... 28 Habitat Use and proximity to water resources ...... 29

vii

Stellenbosch University https://scholar.sun.ac.za

Conclusion ...... 29 References: ...... 31

Chapter Three: The population structure and habitat use of the African elephant (Loxodonta africana) in Majete Wildlife reserve, Malawi ...... 40 Abstract ...... 40 Introduction ...... 41 Methods and Materials ...... 43 Study Area ...... 43 Methods ...... 45 Statistical Analysis ...... 50 Results ...... 50 Discussion...... 56 Conclusion ...... 60 Acknowledgements ...... 61 References ...... 62

Chapter Four: A preliminary investigation of the diet of African elephant (Loxodonta africana) in Majete Wildlife Reserve, Malawi, using stable isotope analysis ...... 67 Abstract ...... 67 Introduction ...... 68 Methods ...... 70 Study Area ...... 70 Faecal and Vegetation Sampling and Analysis ...... 71 Statistical Analysis ...... 73 Results ...... 73 Discussion...... 78 Acknowledgements ...... 83 References ...... 84

Chapter Five: Research findings, conclusions and management recommendations for African Parks Majete management staff ...... 88 Overview ...... 88 Research Findings ...... 88 Conclusions and Management Recommendations ...... 90 Evaluation of management options ...... 92

viii

Stellenbosch University https://scholar.sun.ac.za

Conclusion ...... 96 References ...... 98

Appendix 1 ...... 101 The wildlife reintroduction history of Majete Wildlife Reserve ...... 101

Appendix 2 ...... 102 Appendix 2.1. An example of Majete elephant database ...... 102 Appendix 2.2. An example of a well-known family herd in MWR ...... 103 Appendix 2.3. A table representing the number family herds and bulls in MWR and their total number, sex and age structure in 2015 and 2016 ...... 104

Stellenbosch University https://scholar.sun.ac.za

List of Tables

Table 2.1. African elephant demographic parameters for four different populations collected through individual recognition, radio tracking or culling ……………………………………………………………………..…………21

Table 3.1. Age and sex structure of the elephant population in MWR in 2015 and early 2016. The sex ratios and the proportion of each age class in the population are also shown. Sex for 84 calves and juveniles was not determined………………………………………………………………………………………………………....50

13 15 Table 4.1. δ C and δ N values (‰) of C3 shrubs and trees and C4 grass specimens used as a reference in the stable isotope analysis of the diet of elephant in MWR, Malawi. The standard corrected values were determined by the Isotope Laboratory, Mammal Research Institute, University of Pretoria……..73

Table 4.2. δ15N and δ13C values (‰) of faecal samples representing the diet of elephant in the early dry season 2015, late dry season 2015, the late wet season 2015 and early wet season 2016 in MWR, Malawi. The standard corrected values were determined by laboratory technicians at the Stable Isotope Lab, Mammal Research Institute, University of Pretoria……………………………………………………….75

Table 4.3. Percentage browse consumption of elephant in MWR, Malawi, as determined from stable carbon isotope analysis of faecal samples (n=number of samples). Percentage browse consumption was determined using a Bayesian stable-isotope mixing model using C3 (-29.23‰) and C4 (-13.12‰) plant end-members………………………………………………………………………………………………………………………….77

Stellenbosch University https://scholar.sun.ac.za

List of Figures

Figure 1.1. The location of Malawi on the African continent and the position of Majete Wildlife Reserve in Malawi. The various regions and perennial water sources that occur within the reserve are also shown. (Shapefiles per comms. African Parks (Pty) Ltd.)……………………………………………………………………..4

Figure 3.1. The location of Malawi on the African continent and the position of Majete Wildlife Reserve in Malawi. The various regions and perennial water sources that occur within the reserve are also shown. (Shapefiles per comms. African Parks (Pty) Ltd.)………………………………….……………………………..…44

Figure 3.2. The proportional representation of each age class in the elephant population in MWR, Malawi, in the years 2015/2016……………………………………………………………………………………………………….50

Figure 3.3. The cumulative total number of elephant in MWR since the start of the reintroduction process in 2006. (p=0.00007, r2=0.9971)………………………………………………………………………………….……….51

Figure 3.4. Maps representing elephant presence across MWR in two different wet and dry seasons in relation to altitude and presence of perennial water sources. Number of individuals recorded at each site is correlated to the size of the circular marker…………………………………………………………………….52

Figure 3.5. Map representing the elephant group number and position on MWR in the 2015 aerial survey. (pers comms. African Parks (Pty Ltd.)………………………………………………………………………………….53

Figure 3.6. A joint tree cluster analysis representing the correlation between the presence of various species on MWR………………………………………………………………………………………………………………………………53

Figure 3.7. Maps representing elephants’, more specifically bulls and family herds, water point usage in MWR in the wet and dry seasons of 2014, 2015 and 2016. The size of each pie chart is correlated to the number of observations at each water point……………………………………………………………………………...55

Figure 4.1. The isotopic values of C4 and C3 vegetation, along with the isotopic values of carbon and nitrogen in the diet of elephant (Loxodonta africana) in the early dry season 2015, late dry season 2015, late wet season 2015 and early wet season 2016 in MWR, Malawi…………………………………….76

Stellenbosch University https://scholar.sun.ac.za

List of Abbreviations

AENP Addo Elephant National Park

ANP Amboseli National Park

DNPW Department of National Parks and Wildife, Malawi

KNP Kruger National Park

MWR Majete Wildlife Reserve

SNR Samburu National Reserve

TENP Tsavo East National Park

xii

Stellenbosch University https://scholar.sun.ac.za

Chapter One

Thesis Introduction and Outline

Introduction

Wildlife management is an extremely intricate yet adaptable discipline. Its foundation is based on the combination of scientific research, its practical application and the manipulation of natural systems that all constantly adjust according to the management objectives of management teams for various species and the status of their populations. One such species that requires adaptive forms of management is the African elephant (Loxodonta africana) (Whyte et al. 2003; van Aarde & Jackson 2007). It is one of the most intensively studied species on the African continent, both biologically and ecologically, and this is mainly attributed to the keystone role it plays in the numerous habitats it occupies (Short 1966; Hanks 1969; Laws 1970; Croze 1972; Barnes 1983; Koch et al. 1995; van Aarde et al. 1999; Whitehouse & Schoeman 2002; Shannon et al. 2006; Loarie et al. 2009; Shrader et al. 2012; Kioko et al. 2013; Wittemyer et al. 2013). The elephant is well known for structuring both plant and animal communities and although its effect can be beneficial it can also be detrimental when elephant densities increase in confined areas (Whyte et al. 2003; Skarpe et al. 2004; Western & Maitumo 2004; de Beer et al. 2006). Therefore, the potential impact of the species and its population status is critical to the integrity of the ecosystems it inhabits and so raises many management questions for the protected and non-protected areas it occupies.

The management of the African elephant is further complicated by virtue of the fact that the status of the species varies greatly across its range states as it is more threatened in some countries compared to others (Blanc et al. 2007; Bouche et al. 2011). This is supported by the fact that there have been recent eradications of the species in central and western Africa and yet nearly a third of the entire population occurs in Southern Africa (Blanc et al. 2007). However, only 9% of the entire African continent has been declared as formally protected areas and only 31% of the entire elephant population ranges within these protected areas (Blanc et al. 2007). Coincidently, many elephants still occur outside of protected areas which lends to an additional challenge for wildlife management. In particular, human-elephant conflict runs rife in many parts of Africa, especially when human- agricultural expansion moves into undeveloped land or infringes on protected area boundaries (Hoare & Du Toit 1999; Osborn & Parker 2003; Sitati et al. 2005). Broadly speaking, the primary issues that impede the survival of African elephant populations are habitat loss and fragmentation, human- elephant conflict, poaching for meat and ivory and the negative local impacts of elephant on their

Stellenbosch University https://scholar.sun.ac.za

habitats (Gillson & Lindsay 2003; Osborn & Parker 2003; Stiles 2004; Blanc et al. 2007; van Aarde & Ferreira 2009). These issues are exacerbated by an ever-increasing human population throughout Southern Africa as well as other anthropogenic activities such as economic expansion, poverty and social and environmental human displacement that infringe on remaining wilderness areas (Newmark 2008). Therefore, in order to try and mitigate some of these pressures on the remaining wilderness areas as well as elephant populations there is a great need for clear wildlife management objectives for both protected and unprotected areas throughout the African continent.

In order for wildlife management to minimise the impact of many of these issues faced, not only by the African elephant but many other species as well, an increasing number of reserves, particularly in Southern Africa, are being fenced (Newmark 2008; Hayward & Kerley 2009). Fencing protected areas is beneficial in terms of reducing human wildlife conflict but it essentially increases the isolation of protected areas by creating an artificially closed ecosystem where food, water and space are limited and wildlife is unable to migrate to and from the area (Boone & Hobbs 2004; Hayward & Kerley 2009). Elephants, in particular, are a mixed feeding and water dependent species that naturally shifts its ranges on a seasonal basis (Codron et al. 2006; Chamaille-Jammes et al. 2007; van Aarde & Jackson 2007; Loarie et al. 2009; Codron et al. 2011). The combination of confining an elephant population and preventing its natural seasonal migration, as well as supplementing its resources in the form of artificial waterholes and providing protection from poaching, can rapidly enhance the population growth rate (Boone & Hobbs 2004; van Aarde & Jackson 2007; Shrader et al. 2010). As a result, this increases the foraging pressure of elephant on the vegetation within fenced protected areas that can have large impacts on the woody vegetation structure and its diversity, ultimately causing a loss of vegetation cover and negatively affecting other herbivore species (Duffy et al. 2002; van Aarde & Jackson 2007; Loarie et al. 2009; Mapaure & Moe 2009). Therefore, closed systems such as fenced reserves require wildlife managers to know the status of their elephant populations, whether they are increasing or decreasing and whether numbers should be regulated as the intensive management of the species is fundamental to the mitigation of habitat degradation and additional effects on other biodiversity.

The management of elephants in closed systems is one of the greatest challenges for wildlife management not only because of the potential impact of the species on the biodiversity of the protected area but because to date there is very little understanding of what limits their population numbers (Skarpe et al. 2004; Owen-Smith et al. 2006; van Aarde & Jackson 2007). The African elephant is a long-lived mammal, with a long period of sexual maturity and the longest mammalian gestation period (Moss 2001; Wittemyer et al. 2013). What is understood is that survival of the adults is high,

Stellenbosch University https://scholar.sun.ac.za

the survival of the sub adults (<10 years) is less so and poaching and droughts are two of the main variables that limit elephant numbers (Dudley et al. 2001; Owen-Smith et al. 2005; van Aarde & Jackson 2007). However, the protection and resources provided in protected areas enhance population growth (Boone & Hobbs 2004; Hayward & Kerley 2009). Due to this, many fenced protected areas particularly in Southern Africa, face the challenge of elephant populations rapidly increasing to the point where management needs to implement a form of population regulation such as, translocation, contraception or culling (Cumming et al. 1997; van Aarde et al. 1999; Pimm & van Aarde 2001; Slotow et al. 2005). Therefore, although we do not have a clear understanding of what limits elephant populations in natural systems, wildlife managers have a duty to regulate population numbers due to the artificial nature of fenced protected areas and to ensure the perseverance of other biodiversity in the area.

In order to regulate elephant population numbers in fenced reserves, wildlife managers require an understanding of the variables that contribute towards population growth such as demographics, waterhole use and diet. The diet of elephant, in particular, can be quantified using stable carbon isotope ecology as it addresses the proportion of grass and browse consumed by individuals. Stable carbon isotope ratios are distinct between plants using the C3 photosynthetic pathway (trees, shrubs and forbs) and plants that make use of the C4 photosynthetic pathway (grasses) (Cerling et al. 1999).

Once herbivores consume either C3 or C4 vegetation the carbon from the food is incorporated into the animal body tissues (Codron et al. 2007). Materials such as bone or hair and even the excreted, undigested portion of the diet, the faeces, reflect the herbivore 13C/12C ratios and consequently the proportions of grass and browse consumed (Cerling et al. 1999; Sponheimer et al. 2003; Codron et al. 2005; Codron et al. 2007). Therefore, a greater understanding of such variables will aid in the development of efficient elephant management plans on reserves that will facilitate better population management.

The management conundrum of elephants is so topical for many protected areas throughout Africa, that management organisations such as African Parks (Pty) Ltd. are investigating these very challenges in their own reserves. African Parks (Pty) Ltd., is a non-profit organisation that focuses in developing wildlife parks to be socially and economically viable, especially to the advantage of the local communities. They work in partnerships with governments and local communities and get their primary source of funding from private donors who value the protection and sound management of Africa’s protected areas. The first rehabilitation project taken on by African Parks (Pty) Ltd. was Majete Wildlife Reserve (MWR, hereafter) which is the primary study site for this specific study where African Parks (Pty) Ltd. requested that a population study be conducted on its elephants (Figure 1.1.). MWR

Stellenbosch University https://scholar.sun.ac.za

is located in the lower Shire valley region of southern Malawi (S15o 54’26.6”; E034o44’24.3”) (Figure 1.1.). It is a 700km2 fenced reserve with the northern and eastern boundaries boarded by two perennial rivers, the Mkhulumadzi and the Shire. MWR has two distinct seasons, the wet season that occurs from December to May and the dry season from June to November. The annual precipitation on the reserve is between 680-800mm in the eastern lowlands and 700-1000mm in the western highlands (Wienand 2013). Water availability is affected by season but there are approximately five perennial springs, along with ten artificial waterholes in the reserve.

Figure 1.1. The location of Malawi in the African continent and the position of Majete Wildlife Reserve in Malawi. The various regions and perennial water sources that occur within the reserve are also shown. (Shapefiles per comms. African Parks (Pty) Ltd.)

The vegetation on the reserve is primarily woodland which varies according to altitude. The vegetation types include: high altitude miombo woodland, medium altitude mixed woodland, low altitude mixed woodland and savanna. MWR was gazetted in 1955 but by 2003 most of its large game had been decimated due to poaching and poor management. In 2003, MWR underwent one of Africa’s greatest reintroduction programmes, after a Public Private Partnership (PPP) agreement was made between African Parks, Majete (Pty) Ltd. and the Malawian government. Wildlife reintroductions to MWR were undertaken in stages as the reserve was not yet fenced. The first stage involved fencing a smaller area of 140km2 in the north-eastern region of the reserve, known as the Sanctuary, purely for reintroduction purposes in 2003. In the second stage the remaining boundary of the reserve was fenced and was completed in 2008, after which the sanctuary fence line was removed in 2011 and

Stellenbosch University https://scholar.sun.ac.za

wildlife was able to roam into the greater reserve. In total over 2550 individual animals of 14 different species were reintroduced into MWR. Species reintroduced into the reserve included elephant, black rhino, buffalo, sable, Lichtenstein hartebeest and numerous other antelope species. Details of the reintroduction can be found in Appendix 1.

The reintroduction of wildlife species into an area where they were previously extinct due to anthropogenic pressures, is an effective tool in wildlife management (Muths & Dreitz 2008). The success of a reintroduction program is measured by the successful release of animals followed by their ability to reproduce and form self-sustaining populations in an area (Seddon et al. 2007; Muths & Dreitz 2008). Since the reintroduction a total of 213 elephants into the reserve in 2006, 2008 and 2010, respectively, populations have increased dramatically and apart from periodic aerial surveys (2010, 2012 and 2015), which recently counted 389 individuals in the reserve, there has been no other formal study on the status of the population. Currently, MWR has a no hunting and no culling policy with the intention to relocate surplus animals to restock other protected areas within Malawi. Therefore, it is of high importance to monitor and actively manage the current elephant population on MWR so as to establish a minimum viable population that must remain on the reserve in order to maintain a healthy ecosystem and to determine the number of individuals that will eventually be translocated.

The African elephant population on MWR was selected for this thesis as the population was increasing dramatically which could possibly lead to resource competition with other herbivores. Additionally, the combined biomass of the elephant population could have a detrimental effect of the reserves vegetation if left unchecked. It is important to know how animals utilise natural resources and more specifically to know the minimum and maximum viable populations for an area in order to maintain a healthy mammal population. The findings from this thesis may help to improve other reintroduction programmes and provide a better understanding of how pioneer populations successfully establish themselves in new territories.

Research Question

What is the current population status, habitat use and seasonal diet of Majete Wildlife Reserve’s elephants, five years post reintroduction?

Research Statement

This study presents findings of the demographic changes to MWR’s elephant population and their habitat use and seasonal dietary requirements, five years post reintroduction into the reserve. An assessment was conducted on the population structure, including gender ratio, in 2015/2016.

Stellenbosch University https://scholar.sun.ac.za

Additionally, habitat use of the elephants on MWR was investigated to determine which areas of the reserve they frequented and which water sources were preferred on a seasonal basis. Lastly, seasonal dietary changes were also investigated to determine whether there was a change in the browse and grass content between seasons. Data gathered in this study will aid MWR in the development of sound management strategies that will determine the future course of action in order to maintain and control the elephant population without negatively impacting MWR’s remaining biodiversity.

Hypotheses

Alternative Hypothesis a) Five years post reintroduction, MWRs’ elephant population is steadily increasing: the sex ratio is skewed towards females, calves and juveniles form the largest proportion of the population, whereas the adults form the smallest.

b) Elephants prefer lower altitudinal areas of the reserve in both the dry and wet seasons. In the dry season the elephant’s habitat use is focused around the water sources. However, in the wet season, when water is readily available throughout the reserve in the form of ephemeral pools, elephants increase home ranges and so habitat use is more homogenous and perennial water sources are less frequented.

c) Seasonal diet shifts from mixed feeding (both grass and browse) in the wet season to predominantly browse in the dry season. Null Hypothesis a) Five years post reintroduction MWRs’ elephant population has not dramatically increased, the sex ratio is not skewed and there has been no change in the age structure.

b) Habitat use by elephants does not change on a seasonal basis and their use of perennial water sources is consistent throughout the year.

c) The seasonal diet of the elephants does not shift on a seasonal basis and remains very much the same throughout the year.

Research Goals

The main research goal is to provide MWR with baseline information and guidelines for the development of a sound management plan for the elephant population that currently inhabits the reserve. The guidelines, based on scientific and ecologically sound research, will aim to address the

Stellenbosch University https://scholar.sun.ac.za

future course of action needed to maintain a viable elephant population on the reserve, to preserve spatial heterogeneity and to prevent loss of biodiversity.

Research Objectives and Research Questions

The first objective is to determine how the demographics of MWRs’ elephants has changed five years post reintroduction and to identify whether population changes are associated with environmental variables, such as vegetation and water availability.

a) What is the elephant population size in MWR? b) What is the population’s age/size class structure? (number of males and females, number of adults, juveniles and calves) c) How many elephant herds are there in the reserve? d) What environmental variables in MWR can be associated with the change in population structure?

The second objective is to determine seasonal changes in habitat use, with a focus on which regions of the reserve are regularly frequented by elephants and which perennial water sources are favoured.

a) Does the presence of elephants within MWR differ on a seasonal and yearly basis? b) Which regions of the reserve have the highest and lowest presence of elephants on a seasonal basis? c) Does the use of perennial water sources by elephant change on a seasonal and yearly basis? d) Which perennial water sources are frequented most in the wet and dry seasons?

The third objective is to determine the change in the seasonal diet of MWRs’ elephant population, more specifically to determine the difference in the browse and grass content of the elephant’s diet in the early wet, late wet, early dry and late dry seasons.

a) What is the isotopic composition of C3 and C4 biomass in the elephant’s diets?

b) Is there a seasonal (i.e. early wet, late wet, early dry and late dry) change in the C3 and C4 biomass in the elephant’s diet?

Significance of Research

African Elephants are a keystone species and the potential impact they may have on their habitat and other herbivore species raises management questions in both protected and unprotected areas (Short 1966; Hanks 1969; Laws 1970; Croze 1972; Barnes 1983; Koch et al. 1995; van Aarde et al. 1999;

Stellenbosch University https://scholar.sun.ac.za

Whitehouse & Schoeman 2002; Shannon et al. 2006; Loarie et al. 2009; Shrader et al. 2012; Kioko et al. 2013; Wittemyer et al. 2013). In order to maintain healthy ecosystems it is important to manage any enclosed elephant population, where fencing prevents the seasonal migration of the species (Newmark 2008; Boone & Thompson Hobbs 2004; Hayward & Kerley 2009). The combination of fencing and the installation of artificial waterholes in many of these fenced reserves intensifies the impact of elephants on vegetation due to increased elephant densities and improved foraging opportunities in the drier seasons (Chamaille-Jammes et al. 2007; Shannon et al. 2006). Therefore, it is important for reserve management to have access to all relevant information concerning the status of their elephant populations and the animal’s habitat use in order to compile management plans that will help maintain healthy ecosystems within the protected area.

It is not just the keystone status of the African elephant that reflects the importance of their management. They are a species of high economic value in terms of tourism, where they are featured as a flagship species, and the ivory trade (Gillson & Lindsay 2003; Osborn & Parker 2003; Stiles 2004; Blanc et al. 2007; van Aarde & Ferreira 2009; Wittemeyer et al. 2013). The African elephant was listed in Appendix I in 1997, which banned all international trade of elephants and their parts. The listing was done by CITES (Convention on International Trade in Endangered Species of wild fauna and flora), an organisation that regulates the trade in species and their products (van Aarde & Ferriera 2009; Stiles 2004). The aim of this listing was to attempt to stop the trade in ivory, which was rapidly decreasing elephant numbers in parts of Africa (van Aarde & Ferriera 2009). It was not entirely effective as the number of elephants continued to decline in Central and Western Africa, and increased in Southern and Eastern Africa (Stiles 2004; Blanc et al. 2007). This suggests that there are factors, other than the illegal ivory trade, that can cause elephant population changes such as political and economic factors, government investment in wildlife, the bush meat trade and the commercial use of elephant by-products (Stiles 2004; van Aarde & Ferriera 2009). The status of the African elephant therefore varies across the continent, so much so that it is essential that we aim to conserve as many populations as possible by creating clear management objectives for both protected and unprotected areas by using sound scientific and ecological research.

Lastly, since the reintroduction of elephants into MWR only three ecological studies have been conducted. The first was conducted in 2012 and focused on the effects of elephant browsing on woody trees (Staub et al. 2013). The study found that elephants mostly favoured riparian woodlands and that browsing was negatively related to the distance from water sources (Staub et al. 2013). The second study was conducted in 2013 and investigated changes in the historic woody vegetation cover from 1985 – 2010, as well as elephant water point usage (Wienand 2013). The study found a high loss of

Stellenbosch University https://scholar.sun.ac.za

woody vegetation between 2000 and 2010 which was mainly attributed to the synergistic effects of drought, fire and herbivory (Wienand 2013). Elephant water point use was effected by season, water point altitude and the surrounding vegetation type (Wienand 2013). The third study was conducted in 2013 and investigated the browsing impact of elephant on the woody tree species in MWR (Komoto 2013). The study found that elephants had a negative impact on woody species and the highest impacts occurred in the Sanctuary which is dominated by Acacia/Sclerocarya woodland (Komoto 2013). However, to date there has not been a study that has focused on the elephant population structure and status and how it has responded since the reintroductions. Neither of the two previous studies considered elephant diet specifically and how it changes on a seasonal basis. Therefore, this study, in combination with previous studies conducted on elephants in MWR (Staub et al. 2013; Weinand 2013) will provide the necessary data to compile a long term management plan for the elephants in the reserve. Findings of this research will be used to improve similar reintroduction programmes and provide a better understanding of how elephant populations establish in new environments.

Scope of Limitations

In this project four main limitations were identified and considered.

The first limitation to the research conducted in MWR is that African elephants are a particularly long- lived species (Wittemyer et al. 2013). They have the longest mammalian reproductive life as well as gestation period but have a slow rate of reproduction (Moss 2001; Wittemyer et al. 2013). This presents various problems in research conducted on the population dynamics of African elephants, as there is a large constraint in collecting longitudinal data on the demography of a population (Moss 2001). It could take up to 60 years to determine the complete life history of a herd (Moss 2001). Therefore, due to the fact that the reintroduction of the species into MWR was only completed five years ago and because of time constraints, this study this study was unable to investigate the complete life history and demographics of the population. The study did, however, manage to obtain a cross section of the population status of the elephants for the years 2015/2016, thus providing important baseline data that can be used for future research.

The second limitation to this study was the condition of the roads. The road network in the reserve is fairly limited and is only passable by 4x4 vehicles in the majority of the park. For this reason little time was spent in certain sections of the reserve and daily or even weekly transects of all the roads in the reserve were impossible to conduct. Additionally, in the wet season in particular, the vegetation on the reserve is mainly dense woodland with a few small savanna areas that are located in the south.

Stellenbosch University https://scholar.sun.ac.za

Visibility was thus reduced and this combined with the fact that the majority of the elephants in the reserve are not habituated towards cars or humans made it difficult to collect demographic data in certain areas. In order to overcome this limitation the project made use of camera traps that were placed at artificial waterholes throughout the reserve as well as at a few natural springs. Therefore, data could still be collected in a non-invasive manner and on a continuous basis.

The third limitation to the study was that there was little or no demographic data available from when the elephants were first reintroduced into the reserve in 2006, 2008 and 2010 except for total numbers. So results could not be compared with regards to how the population has changed over time, specifically referring to sex ratios and age groups.

The fourth limitation to the study was that the elephant faeces collected for the seasonal diet study was only collected from main road networks and around artificial waterholes on the reserve. Although the road network is limited with little access to the interior areas of MWR it does have access to all vegetation types. Therefore, in order to ensure that all possible habitat types are included, it is imperative that sampling of elephant faeces is conducted throughout the reserve.

Assumptions

a) Individuals that were captured on camera in a continuous sequence of photographs in a short time period, between 10 to 60 min, were assumed to be from the same herd.

b) Once an extended period of time, longer than 60 min, had passed between photographs, individuals we assumed to be from a different herd.

c) Bulls of the age categories, small, medium and large adults, were assumed to have left their original family herds (size classes were derived from the following literature: Moss 1996; Whitehouse & Hall-Martin 2000; Moss 2001). Bulls tend to leave their family herds between the ages of 11 and 13 years (Moss 1996; Whitehouse & Hall-Martin 2000; Moss 2001). If an individual was captured on photograph with the same family herd multiple times, the individual was assumed to still belong to the herd. However, if an individual male estimated to be a small adult appeared alone or with different family herds in photographs, notes were taken but the individual was assumed to range alone.

d) If more than one family herd was recorded while feeding or at waterholes, notes were taken as separate herds may be part of a larger family group and so could possibly be related.

Stellenbosch University https://scholar.sun.ac.za

Brief Chapter Overview

This master’s thesis is composed of five chapters. Chapter One is a basic introduction to the subject matter by providing some background information about the importance of sound elephant management in protected areas, the research question and statement, the hypotheses, the goals of the study, the objectives, significance of the research and the assumptions and limitations of the study. Chapters Two, Three and Four have been compiled as stand-alone manuscripts to enable publication in peer-reviewed journals. Due to this there is some repetition between chapters. Chapter Five serves as both a discussion chapter for the thesis as well as a management recommendation document for African Parks Majete (Pty) Ltd. Therefore, the literature review is not discussed in Chapter Five.

Chapter Two presents a literature review that provides comprehensive background information on the influence of population demographics, resource use and range distribution on the management of the African elephant in protected areas.

Chapter Three describes how the population structure and demographics of MWRs’ elephants has changed five years post reintroduction onto the reserve. The population age structure and sex ratio were determined for 2015/2016 and the changes since reintroduction were discussed. Additionally, it was determined whether there was a seasonal change in habitat use by the elephants and this was discussed in terms of presence/absence around the reserve and perennial water source usage.

Chapter Four describes the change in the seasonal diet of MWRs’ elephant population. The effects of the different seasons (early wet, late wet, early dry and late dry) on elephant diet was analysed using the C3 and C4 values in elephant faeces and thus the difference in the browse and graze content of the diet was observed. The results and their implications towards elephant management were then discussed and compared to other dietary studies conducted on the African elephant.

Chapter Five summarises the main research findings of this thesis in terms of population structure and diet and proposes recommendations for the future management of MWR’s elephants.

Stellenbosch University https://scholar.sun.ac.za

References

 Barnes, R.F.W. (1983) The Elephant Problem in Rusha National Park, Tanzania. Biological Conservation, 26, 127-148.  Blanc, J.J., Barnes, R.F.W., Craig, G.C., Dublin, H.T., Thouless, C.R., Douglas-Hamilton, I. & Hart, J.A. (2007) African Elephant Status Report 2007: and update from the African Elephant Database. Occasional Paper Series of the IUCN Species Survival Commission, No 33. IUCN/SSC African Elephant Specialist Group IUCN, Gland, Switzerland. Vi +276 pp.  Boone, R.B. & Hobbs, T. (2004) Lines around fragments: effects of fencing on large herbivores. African Journal of Range and Forage Science, 21(3), 147-156.  Bouche, P., Douglas-Hamilton, I., Wittemyer, G., Nianogo, A.J., Doucet, J., Lejeune, P. & Vermeulen, C. (2011) Will elephants soon disappear from West African savannahs?. Plos One, 6(6), 1-11.  Cerling, T.E., Harris, J.M. & Leakey, M.G. (1999). Browsing and Grazing in Elephants: The Isotope record of Modern and Fossil Proboscideans. Oecologia, 120(3), 364–374.  Chamaille-Jammes, S., Valeix, M. & Fritz, H. (2007) Managing heterogeneity in elephant distribution: interactions between elephant population density and surface water availability. Journal of Applied Ecology, 44, 625-633.  Codron, D., Codron, J., Lee-Thorp, J.A., Sponheimer, M., De Ruiter, D., Sealy, J., Grant, R. & Fourie, N. (2007) Diets of savanna ungulates from stable carbon isotope composition of faeces. Journal of Zoology, 273(1), 21-29.  Codron, D., Codron, J., Sponheimer, M., Lee-Thorp, J.A., Robinson, T., Grant, C.C. & de Ruiter, D. (2005) Assessing diet in savanna herbivores using stable carbon isotope ratios of faeces. Koedoe, 48, 115-124.  Codron, J., Codron, D., Lee-Thorp, J.A., Sponheimer, M., Kirkman, K., Duffy, K.J. & Sealy, J. (2011) Landscape-scale feeding patterns of African elephants inferred from carbon isotope analysis of feces. Oecologia, 165, 89-99.  Codron, J., Lee-Thorp, J.A., Sponheimer, M., Codron, D., Grant, R.C. & De Ruiter, D.J. (2006) Elephant (Loxodonta africana) Diets in Kruger National Park, South Africa: Spatial and Landscape Differences. Journal of Mammalogy, 87(1), 27-34.  Croze, H. (1972) A modified photogrammetric technique for assessing age-structures of elephant populations and its use in Kidepo National Park. East African Wildlife Journal, 10, 91- 115.

Stellenbosch University https://scholar.sun.ac.za

 Cumming, D.H.M., Fenton, M.B., Rautenbach, I.L., Taylor, R.D., Cumming, G.S., Cumming, M.S., Dunlop, J.M., Ford, G.S., Hovorka, M.D., Johnston, D.S., Kalcounis, M.C., Mahlanga, Z., and C.V. Portfors. (1997) Elephants, woodlands and biodiversity in miombo woodland in Southern Africa. South African Journal of Science, 93, 231-236.  de Beer, Y. Kilian, W., Versfeld, W. & van Aarde, R.J. (2006) Elephants and low rainfall alter woody vegetation in Etosha National Park, Namibia. Journal of Arid Environment, 64, 412-421.  Dudley, J.P., Craig, G.C., Gibson, D.St., Haynes, G. & Kilmowicz, J. (2001) Drought mortality of bush elephants in Hwange National Park, Zimbabwe. African Journal of Ecology, 39, 187-195.  Duffy, K.j., van Os, R., Vos, S., van Aarde, J. Ellish, G. & Stretch, A-M.B. (2002) Estimating impact of reintroduced elephant on trees in a small reserve. South African Journal of Wildlife Research, 32(1), 23-29.  Gillson, L. & Lindsay, K. (2003) Ivory and ecology – changing perspectives on elephant management and the international trade in ivory. Environmental Science, 6, 411-419.  Hanks, J. (1969) Seasonal breeding of the African elephant in Zambia. African Journal of Ecology, 7(1), 167-180.  Hayward, M.W. & Kerley, G.I.H. (2009) Fencing for conservation: Restriction of evolutionary potential or a riposte to threatening processes? Biological Conservation, 142, 1-13.  Hoare, R.E. & Du Toit, J.T. (1999) Coexistence between People and Elephants in African Savannas. Conservation Biology, 13(3), 633-639.  Kioko, J., Zink, E., Sawdy, M. & Kiffner, C. (2013) Elephant (Loxodonta africana) demography and behaviour in the Tarangire-Manyara Ecosystem, Tanzania. South African Journal of Wildlife Research, 43(1), 44-51.  Koch, P.L., Heisinger, J., Moss, C., Carlson, R.W., Fogel, M.L. & Behrensmeyer. (1995) Isotopic Tracking of Change in Diet and Habitat Use in African Elephants. Science, 267, 1340-1343.  Komoto, S.C. (2013) Post translocation browsing impact of African elephant on woody tree species in Majete Wildlife Reserve, Malawi. MSc Thesis, University of Malawi, Malawi.  Laws, R.M. (1970) Elephants as agents of habitat and landscape change in East Africa. Oikos, 21(1), 1-15.  Loarie, S.R.,van Aarde, R.J. & Pimm, S.L. (2009) Fences and artificial water affect African savannah elephant movement patterns. Biological Conservation, 142(12), 3086-3098.  Mapaure, I. & Moe, S.R. (2009) Changes in the structure and composition of miombo woodlands mediated by elephants (Loxodonta africana) and fire over a 26-year period in north-western Zimbabwe. African Journal Ecology, 47, 175-183.

Stellenbosch University https://scholar.sun.ac.za

 Moss, C. (1996) Getting to know a population. Studying Elephants. (eds K. Kangwana), pp: 58- 74. African Wildlife Foundation. Kenya.  Moss, C.J. (2001) The demography of an African elephant (Loxodonta africana) population in Amboseli, Kenya. Journal of Zoology, 255, 145-156.  Muths, E. & Dreitz, V.J. (2008) Monitoring programs to assess reintroduction efforts: a critical component in recovery. Animal Biodiversity and Conservation, 31, 47-56.  Newmark, W.D. (2008) Isolation of African Protected Areas. Frontiers in Ecology and the Environment, 6(6), 321-328.  Osborn, F.V. & Parker, G.E. (2003) Towards an integrated approach for reducing the conflict between elephants and people: a review of current research. Oryx, 37(1), 80-84.  Owen-Smith, N., Kerley, G.I.H., Page, B., Slotow, R. & van Aarde, R.J. (2006) A scientific perspective on the management of elephants in the Kruger National Park and elsewhere. South African Journal of Science, 102, 389-394.  Owen-Smith, N., Mason, D.R. & Ogutu, J.O. (2005) Correlates of survival rates for 10 African ungulate populations: density, rainfall and predation. Journal of Animal Ecology, 74, 774-788.  Pimm, S.L. & van Aarde, R.J. (2001) African elephants and contraception. Nature, 411, 766- 767.  Seddon, P.J., Armstrong, D.P. & Maloney, R.F. (2007) Developing the Science of Reintroduction Biology. Conservation Biology, 21 (2), 303-312.  Shannon, G., Page, B., Slotow, R. & Duffy, K. (2006) African elephant home range and habitat selection in Pongola Game Reserve, South Africa. African Zoology, 41 (1), 37-44.  Short, R.V. (1966) Oestrous behaviour, ovulation and the formation of the corpus luteum in the African elephant, Loxodonta africana. African Journal of Ecology, 4(1), 56-68.  Shrader, A.M., Bell, C., Bertolli, L. & Ward, D. (2012) Forest of the trees: At what scale do elephants make foraging decisions? Acta Oecologica, 42, 3-10.  Shrader, A.M., Pimm, S.L, & van Aarde, R.J. (2010) Elephant survival, rainfall and the confounding effects of water provision and fences. Biodiversity Conservation, 19, 2235-2245.  Sitati, N.W., Walpole, M.J. & Leader-Williams, N. (2005) Factors affecting the susceptibility of farms to crop raiding by African Elephants: using a predictive model to mitigate conflict. Journal of Applied Ecology, 42, 1175-1182.  Skarpe, C., Aarrestad, P.A., Andreassen, H.P., Dhillion, S.S., Dimakatso, T., du Toit, J.T., Halley, D.J., Hytteborn, H., Makhabu, S., Mari, M., Marokane, W., Masunga, G., Modise, D., Moe, S.R., Mojaphoko, R., Mosugelo, D., Motsumi, M., Neo-Mahupeleng, G., Ramotadima, M., Rutina, L., Sechele, L. & Sejoe, T.B., Stokke, S., Swenson, J.E., Taolo, C., Vandewalle, M. & Wegge, P.

Stellenbosch University https://scholar.sun.ac.za

(2004) The return of the giants: Ecological Effects of an Increasing Elephant Population. Ambio, 33(6), 276-282.  Slotow, R., Garai, M.E., Reilly, B., Page, B.& Carr, R.D. (2005) Population dynamics of elephants re-introduced to small fenced reserves in South Africa. South African Journal of Wildlife Research, 35(1), 23-32.  Sponheimer, M., Robinson, T., Ayliffe, L., Passey, B., Roeder, B., Shipley, L., Lopez, E., Cerling, T., Dearing, D. & Ehleringer, J. (2003) An experimental study of carbon-isotope fractionation between diet, hair, and faeces of mammalian herbivores. Canadian Journal of Zoology, 81, 871-876.  Staub, C.G., Binford, M.W. & Stevens F.R. (2013) Elephant herbivory in Majete Wildlife Reserve, Malawi. African Journal of Ecology, 51(4), 536-543.  Stiles, D. (2004) The ivory trade and elephant conservation. Environmental Conservation, 31(4), 309-321.  van Aarde, R.J. & Ferreira, S.M. (2009) Elephant populations and CITES trade resolutions. Environmental Conservation, 36(1), 8-10.  van Aarde, R.J. & Jackson, T.P. (2007) Megaparks for metapopulations: Addressing the causes of locally high elephant numbers in Southern Africa. Biological Conservation, 134, 289-297.  van Aarde, R.J., Whyte, I. & Pimm, S. (1999) Culling and the dynamics of the Kruger National Park African elephant population. Animal Conservation, 2, 287-294.  Western, D. & Maitumo, D. (2004) Woodland loss and restoration in a savanna park: a 20 year experiment. African Journal of Ecology, 42, 111-121.  Whitehouse, A.M. & Hall-Martin, A.J. (2000) Elephants in Addo Elephant National Park, South Africa: reconstruction of the population’s history, Oryx, 34(1), 46-55.  Whitehouse, A.M. & Schoeman, D.S. (2002) Ranging behaviour of elephants within a small fenced area in Addo Elephant National Park, South Africa. African Zoology, 38(1), 95-108.  Whyte, I.J., van Aarde, R.J. and Pimm, S.L. (2003) Kruger’s Elephant Population: Its size and Consequences for Ecosystem Heterogeneity. The Kruger Experience. (eds J.T. du Toit, K.H. Rogers & H.C. Biggs), pp: 332-348. Island Press. Washington.  Wienand, J. J. (2013) Woody vegetation change and elephant water point use in Majete Wildlife Reserve: implications for water management strategies. MSc Thesis, University of Stellenbosch, South Africa.  Wittemyer, G., Daballen, D. and Douglas-Hamilton, I. (2013) Comparative Demography of an At-Risk African Elephant Population. Plos One, 8(1), e53726.

Stellenbosch University https://scholar.sun.ac.za

Chapter 2

African elephant (Loxodonta africana) population demographics and resource use in protected areas: A review

F.A. Forrer 1, A. Leslie1

1Department of Conservation Ecology and Entomology, Stellenbosch University, Matieland, Western Cape, 7602, South Africa

Abstract

African elephants (Loxodonta africana) are facing a continental wide decrease in numbers as populations are becoming increasingly isolated in remaining wilderness areas. Many protected areas, particularly in Southern Africa, are however experiencing an increase in elephant population growth rates. This has mainly been credited to two factors: fencing and water supplementation. Fencing limits elephant’s migration and intensifies the species’ impact on the vegetation in an area, potentially resulting in woody species degradation or loss. Water supplementation, concurrently, affords elephants the opportunity to expand their dry season ranging, allowing them to access areas that would normally be unreachable due to lack of water. The high population growth rates caused by the combination of these two factors necessitates that elephant numbers be carefully managed in these protected areas in order to maintain biodiversity and ecosystem functions. Therefore, it is essential that reserve managers understand the primary factors that affect demographic vital rates of elephants and how these can be manipulated to control population growth. In addition, it is important to understand the species’ basic nutritional requirements/diet and investigate their movement and habitat use in an area in order to implement effective management strategies. In this review, research findings of peer-reviewed literature studying elephant population demographics, habitat use and diet were synthesised. The review discussed two core areas i) Elephant demographic vital rates and population growth and ii) Elephant resource use and range distribution in protected areas.

Key Words: African Elephants, demographic vital rates, diet, habitat use, Loxodonta africana.

Stellenbosch University https://scholar.sun.ac.za

Introduction

The African elephant (Loxodonta africana), a large pachyderm, ranges over 22% of the African continent but its status varies significantly across its range states as the long term survival of some national populations are more threatened than others (Blanc et al. 2007). Overall, total African elephant population numbers are declining and in the last century this was attributed to the commercial pursuit of ivory, safari hunting and the expansion of human populations (Milner-Gulland & Beddington 1993; Hoare & du Toit 1999; Skarpe et al. 2004). As a result CITES (the Convention on International Trade in Endangered Species of Wild Fauna and Flora) has given the species dual status (Stiles 2004; Lovett 2009; Jones 2016). The majority of elephant populations throughout Africa are listed under Appendix I, banning international trade (Stiles 2004; Lovett 2009; Jones 2016). However, in Botswana, Namibia, South Africa and Zimbabwe elephants are listed under Appendix II, permitting international trade due to well managed and generally increasing population numbers (Stiles 2004; Lovett 2009; Jones 2016). There are three subspecies of elephant on the African continent: the savanna elephant (Loxodonta africana africana) which occurs in Eastern and Southern Africa, the forest elephant (Loxodonta africana cyclotis) which primarily occurs in the Congo basin and Central Africa and lastly, the West African elephant which occurs in both forest and savanna systems but are taxonomically uncertain (Blake & Hedges 2004; Blanc et al. 2007). In the interest of this review, only the savanna elephants (Loxodonta africana africana) will be discussed.

Southern Africa accounts for 39% of the savannah elephant’s range area and contains the largest known number of elephants relative to any other region on the continent (Blanc et al. 2007). Despite this, elephant populations have become progressively fragmented as their habitat range has decreased in response to an increasing human population (Newmark 2008; Hayward & Kerley 2009). In sub Saharan Africa the human population more than doubled from 1975-2001, which has led to an increase in the development of rural settlements, subsistence agriculture, deforestation and unsustainable hunting of wildlife (Newmark 2008; Hayward & Kerley 2009). The isolation and fragmentation of wilderness areas has progressed to the point where governmental or private protected areas are all that is left of the once vast expanses of undeveloped land (Newmark 2008; Hayward & Kerley 2009). Additional factors that promoted the isolation of protected areas include: fences, roads, over hunting and disease (Newmark 2008). Fencing, in particular, prevents the migration of elephants into and out of reserves, limiting elephants from accessing patches of higher quality food beyond protected area boundaries, but drastically reducing human-wildlife conflict (Osborn & Parker 2003; Boone & Hobbs 2004; Sitati et al. 2005; Guldemond & Van Aarde 2008, Loarie et al. 2009a).

Stellenbosch University https://scholar.sun.ac.za

Elephants are mixed feeders and their large body size and robust feeding allows them to feed on a variety of plants and specific plant parts such as bark, fruit and bulbs (Kerley & Landman 2006; Owen- Smith & Chapota 2012). In Addo Elephant National Park (AENP), South Africa, elephants fed on 146 different plant species throughout the year and in Chobe National Park, Botswana, 94% of the elephant’s dry season diet was composed of stems, bark and roots from various plant species (Lombard et al. 2001; Kerley & Landman 2006; Owen-Smith & Chapota 2012). Additionally, elephants are a water dependent species and so need to remain within a few kilometres of water sources in the dry season (Stokke & du Toit 2002; Chamille-Jammes et al. 2008). The annual rainfall across the African continent strongly correlates with the historic distribution of elephants in terms of occurrence (de Boer et al. 2013). The species would migrate on a seasonal basis in search of resources that would reduce foraging pressure on any one particular area (de Boer et al. 2013). Presently, elephants are unable to migrate as they once did and so with water supplementation in many protected areas, elephant numbers and vegetation damage increase, promoting the principle known as the ‘elephant problem’ (Barnes 1983; Whyte et al. 2003; van Aarde & Jackson 2007).

Elephants are a keystone species, as they structure both plant and animal communities (Caughley 1976). The fencing of reserves results in elephants having large impacts on woody vegetation (Caughley 1976; Cumming et al. 1997; Boone & Hobbs 2004). The ‘elephant problem’ is where the vegetation/habitat of an area changes due to high elephant densities (Napier Bax & Sheldrick 1963; Duffy et al. 2002; Cerling et al. 2004; Guldemond & van Aarde 2007; van Aarde & Jackson 2007). Evidence of this was presented in Cumming et al. (1997) who found that an increasing elephant population led to a rise in tree felling in Zimbabwe. Additionally, Mapaure and Moe (2009) found that the presence of elephants altered the composition of miombo woodland, decreasing abundances of high preference species such as Brachystegia boehmii. The ‘elephant problem’ is not only limited to fencing but the combination of fencing and the increasing complexity of the surrounding matrix of protected areas prevents elephant populations from migrating across landscapes, resulting in high population numbers in protected areas and subsequent vegetation damage (Boone & Hobbs 2004, Van Aarde & Jackson 2007).

There may be a continental wide decrease in elephant numbers but well protected areas, particularly in Southern Africa, are experiencing an increase in elephant population growth rates (Slotow et al. 2005; Blanc et al. 2007). This can be credited to an increasing number of such areas being fenced as well as the supplementation of resources in the form of artificial waterholes (Chamaille-Jammes et al. 2007). Reserve managers face the challenge of managing high elephant population densities while attempting to maintain biodiversity and ecosystem function (Chamaille-Jammes et al. 2007; Loarie et al. 2009b; Gillson & Lindsay 2003). Thus, the knowledge of mechanisms that limit or regulate large

Stellenbosch University https://scholar.sun.ac.za

herbivore populations is important, not only to theoretical understanding of populations, community and ecological patterns but also to our ability to effectively maintain assemblages of these organisms and the ecological processes they facilitate. This review synthesizes the findings of peer-review literature on elephant populations and focuses on two core areas i) Elephant demographic vital rates and population growth and ii) Elephant resource use and range distribution in protected areas.

Elephant Demographic Vital Rates and Population Growth

The central question behind the theory of population dynamics is to identify the stabilising mechanisms of species vital rates that have the potential to lead to population regulation (Murdoch 1994). The vital rates of a species are defined as the rates of survival, development and reproduction and they usually depend on the age, size and development stages of individuals within populations (Neubert & Caswell 2000). In order to estimate the vital rates of a population the relative contribution of each vital rate to the variability in population growth, the elasticity, as well as the amount of temporal variability in each vital rate, the sensitivity, needs to be addressed (de Kroon et al. 2000; Tuljapurkar et al. 2003). The elasticity of African elephants is fairly well understood and developed in the form of the life history traits of the species (Moss 2001). They are characteristic of large ungulates, being strongly iteroparous with low annual fecundity and high annual adult survivorship; they produce only one offspring per reproductive bout and have long gestation periods (de Kroon et al. 2000; Moss 2001; Tuljapurkar et al. 2003; Wittemyer et al. 2007b). However, elephants also exhibit unique life history traits, namely sensitivity of vital rates, that improves their ability to respond to temporal variability in ecological conditions or other factors, such as poaching, governmental corruption and conflict that influence population demographic fluctuations (Osborn & Parker 2003; Boone & Hobbs 2004; Guldemond et al. 2005; Guldemond & Van Aarde 2008; Wittemyer et al. 2007b; Chamaille- Jammes et al. 2008; de Boer et al. 2013). These unique traits include, the longest gestation period of all terrestrial mammals (22 months), an extended period of parental care of young in which weaning typically occurs over 2 years, an overlap of dependant offspring and lastly elephants necessitate extensive, long term energy investment in offspring (Whitehouse & Hall-Martin 2000; Moss 2001; Wittemyer et al. 2007b; Wittemyer et al. 2013). Therefore, the elasticity or the analysis of life history traits of long lived herbivores, such as elephant, indicate that the population growth rate is most sensitive to the change in adult female survival and the fecundity of prime age individuals, whereas the sensitivity of the population to temporal variation is most affected by juvenile survival (de Kroon et al. 2000; Gaillard et al. 2000; Eberhardt 2002; Tuljapurkar et al. 2003).

The elasticity and the sensitivity of an elephant population are interdependent and both are strongly influenced by environmental and anthropogenic factors. The reproductive phenology of the African

Stellenbosch University https://scholar.sun.ac.za

elephant is a prime example of the interdependent relationship between the elasticity and sensitivity of a population as it is influenced by both ecological factors as well as life history constraints (Clutton- Brock 2002). As a consequence of their 22 month gestation period, elephants optimise the timing onset of reproduction so as to maximise the use of known and future resources to coincide with the energetic investment in reproduction (Wittemyer et al. 2007b; Trimble et al. 2009; Shrader et al. 2010). The initiation of the female reproductive bout is dependent on the conditions during the season of conception but is timed so parturition occurs during the most likely period of high vegetation primary productivity 22 months later (Gough & Kerley 2006; Wittemyer et al. 2007b). More specifically, there is an ecological control on elephant reproductive rates as the onset of the rainy season, when primary productivity and food quality increase, triggers an increase in progestin hormone levels and so the onset of reproductive activity (Wittemyer et al. 2007b). Therefore, the progestin levels in pregnant and non-pregnant females is closely correlated with the quality of available food resources and are impacted by age, decreasing the older the individual is, and inter- pregnancy interval (Wittemyer et al. 2007b). This gives a physiological basis for condition dependant oestrus and its influence on general population demographic fluctuations in reaction to climate variability.

The African elephant is a species of high conservation concern and so a clear understanding of any population’s elasticity and sensitivity is essential to preserve the integrity of the ecosystems elephants inhabit and for protected area management. The majority of the existing data on elephant demographic parameters has been compiled from well protected populations that were stable or increasing at the time of the study (Moss 2001; Gough & Kerley 2006). However, in order to determine how the vital rates of different elephant populations vary it is necessary to discuss both populations that are well protected and those that are less protected and face pressures such as poaching. The following section will discuss the primary demographic vital rates that have the greatest influence on population growth. These include calving interval, average age of females at first calving, annual mortality, and density in order to observe the differences between four existing elephant populations with varying population histories, two from fenced reserves (Kruger National Park (KNP) and Addo Elephant National Park (AENP)) and two from unfenced reserves (Amboseli National Park (ANP) and Samburu National Reserve (SNP) (Table 2.1.).

Stellenbosch University https://scholar.sun.ac.za

Table 2.1. African elephant demographic parameters for four different populations collected through individual recognition, radio tracking or culling.

Annual Calving Annual Yearly Protection Age of first Density (per population Location Interval Mortality Rainfall References Status calf (years) km2) growth rate (Years) (%) (mm) (%)

Addo N.P. Gough & Fenced 3.3 12.5 1.4 4 5.8 392 South Africa Kerley 2006

Whyte et al. Kruger N.P 1998; Fenced 3.8 13 3.2 0.52 - 450-700 South Africa Freedman et al. 2009

Amboseli Unfenced 4.5 12 4.15 0.36 2.2 341 Moss 2001 N.P. Kenya

Wittemyer Samburu et al. 2005; Unfenced 4.0 11.34 4.7 0.27 0.17 350 N.R. Kenya Wittemyer et al. 2013

Mean calving interval is considered the single most important factor that influences the population growth rate of elephants and it is dependent on the density of the population as well as the resources available (Moss 2001). The mean calving interval of African elephants can range from 2.9 years to 9.1 years, as derived from culled samples from several different populations throughout Africa (Eltringham 1982, as cited in Moss 2001; Freeman et al. 2008). The length of average calving intervals will increase in resource limited, high density or nutritionally stressed populations and this has the potential to slow population growth as individuals give birth less frequently (Wittemyer et al. 2013). The mean calving interval for the unfenced protected areas of ANP and SNR is 4.5 years and 4.0 years respectively (Table 2.1.). These mean calving intervals are longer than both fenced reserves, AENP and the KNP (Table 2.1.). The rainfall at the time these studies were undertaken was higher in both fenced reserves compared to the unfenced reserves (Table 2.1.), indicating that the populations of SNR and ANP were stressed and the pressures of poaching, resource competition with surrounding human settlements and limited rainfall increased the calving interval of these populations (Moss 2001; Wittemyer et al. 2005; Wittemyer et al. 2013). The mean calving interval of populations is, however, is a result of a number of variables and in fenced reserves, where resources are supplemented and

Stellenbosch University https://scholar.sun.ac.za

human-elephant conflict is minimal, the mean calving interval is likely to be an outcome of high population density.

Density dependent effects is a management approach that has the potential to limit elephant populations in protected areas without resorting to extreme measures such as culling (Laws 1970a; van Aarde et al. 1999). In general, the evidence for density dependence in large mammal populations, such as elephant, indicates that the vital rates, mainly fecundity and mortality, change according to food resources available (Fowler 1987, as cited in Gough and Kerley 2006). More specifically, as population density increases to its highest levels there is a series of changes in demographic vital rates of populations. Firstly, an increase in juvenile mortality rate; secondly, an increase in the age of first reproduction; thirdly, a decrease in adult female reproductive rate and lastly, an increase in adult mortality rate (Gaillard et al. 2000; Eberhardt 2002; Gough & Kerley 2006). The reason as to why juvenile survival rate is more sensitive to density dependence compared to adult survival rate is because adult survival rate has a higher elasticity and is buffered by temporal variation (Gaillard et al. 1998; Gaillard et al. 2000; Gough & Kerley 2006). These changes in demographic vital rates of populations of large mammals have been recorded in a number of different species including red deer, Cervus elephus (Clutton-Brock et al. 1987).

Very little density dependence effects were recorded in the ANP elephant population. It is the only population that has been studied for the extended period of time of 27 years (Moss 2001; Wittemyer et al. 2013). The elephant population has not been disrupted in terms of age and sex structure because there has been no poaching for ivory, no culling and no fencing restricting movements (Moss 2001). Thus, the slower reproduction rate (calving interval) displayed in ANP when compared to SNR and KNP is due to the older age structure of the population and not density (Moss 2001; Wittemyer et al. 2013). Similarly, no density dependence was found in AENP, the population periodically reached high densities but the impacts were mitigated through the expansion of the park and resource provisioning that dampened the ecological processes that influence elephant populations (Whitehouse & Hall- Martin 2000; Gough & Kerley 2006; Wittemyer et al. 2013). Additionally, the vegetation type known as “thicket” is evergreen and highly nutritious and is able to support higher densities of elephants in smaller areas compared to the savanna found in KNP, ANP and SNR (Whitehouse & Hall-Martin 2000; Gough & Kerley 2006).

The average age of a female when she has her first calf or the average primiparous age, is another important factor that can influence a population’s growth rate (Whyte et al. 1998; Moss 2001). The younger the primiparous age of a female elephant the longer her reproductive life span, and as female elephants remain fertile well into their fifties, they have the longest reproductive life span of all

Stellenbosch University https://scholar.sun.ac.za

mammals, therefore greatly contributing towards population growth (Moss 2001). Modelling the KNP population shows that if about 300 females are culled or sterilised just prior to their first pregnancy the population could be driven to extinction (Whyte et al. 1998). The average age of first calving was 12 years for ANP and 11.34 years for SNR (Table 2.1.). Whereas the average age of first calving was 13 years for the KNP and 12.5 years for AENP (Table 2.1.). Wittemyer et al. (2013), suggest that the age of first reproduction is positively correlated with rainfall and density (t=2.80, p=0.021). Similarly, Shrader et al. (2010) found that elephants born in high rainfall years were most likely to survive compared to elephants born in low rainfall years (Rainfall in year of birth: t=2.50, p=0.014). Table 2.1. shows that despite the high density of elephants in AENP the age of first calving is only marginally higher than SNR and ANP but not as high as KNP. This indicates that the variation in the age of first reproduction is a complex function of a female’s early growth, nutrition, individual variation in condition and size when close to average age of reproduction, social and ecological conditions at the time and not just density and rainfall (Moss 2001). This is a life history trait that trades off between starting reproduction and maintaining individual condition and growth and is reflected in the observation that the calves of youngest females had a 50% probability of mortality whereas calves of older females had a 24% probability of mortality (Moss 2001).

The annual mortality within populations is a reflection of resource availability, age structure of a population as well as anthropogenic factors such as poaching. The elephant population at SNR was subject to consistent human pressure and predation throughout the study period and so the life expectancy of males (18.9 years) and females (21.8 years) (Wittemyer et al. 2005; Wittemyer et al. 2013) was much lower than the life expectancy of males (24 years) and females (40 years) at ANP (Moss 2001) (Table 2.1.). Furthermore, the reproductive effort of the elephants in SNR was greater than ANP as they had an earlier age of first reproduction and shorter calving interval, this is similar to other populations with human impacts or that are recovering (Lewis 1984; Moss 2001; Kioko et al. 2013; Wittemyer et al. 2013). Similar to the SNR population, the elephants of the KNP once suffered from high human impacts in the form of culling, the population has since recovered through great reproductive effort in the form of an even lower calving interval and a lower age of first calf (Whyte et al. 1998). It is therefore evident that past management methods of protected areas, location, political and environmental factors influenced the demographic vital rates of elephant populations (Lewis 1984; Moss 2001; de Boer et al. 2013; Kioko et al. 2013; Wittemyer et al. 2013).

Elephant resource use and range distribution in protected areas

Animal population ecology is strongly influenced by variation in the temporal and spatial quality of resources and their abundance (Brown 1984). The seasonal changes in resources, such as plant species

Stellenbosch University https://scholar.sun.ac.za

composition and abundance, can have a strong effect on herbivore diet, essentially effecting the community dynamics and life history strategies of species (Ostfeld & Keesing 2000; Yang et al. 2008). The African elephant is a water-dependent species and so its movement and presence in an area is strongly influenced by the availability and quality of surface water sources (de Boer et al. 2000; Chamaille-Jammes et al. 2007; Martin et al. 2010). African elephants are also classified as mixed feeders as they consume large amounts of grass and browse (Wing & Buss 1970; van der Merwe et al. 1988). At great densities the species has been known to modify woody vegetation in protected areas (Wing & Buss 1970; van der Merwe et al. 1988; Jacobs & Biggs 2002; O’Connor et al. 2007; Mapaure & Moe 2009). The ratio of browse: grass consumed by elephants varies widely across seasons and habitats, some elephant populations have been classified as near pure browsers whereas others have been classified as near pure grazers (Wing & Buss 1970; Guy 1976; van der Merwe et al. 1988). Therefore, elephant populations and their distribution are strongly influenced by the resources available to them.

Rainfall patterns not only drive the availability of natural surface water but also the vegetation productivity in an area and both these resources affect the distribution and densities of elephant populations (Loarie et al. 2009b; Martin et al. 2010; de Boer et al. 2013). The diet of elephants is influenced by the species’ large body size which requires them to bulk feed on many different trees, shrubs and grasses (Clauss et al. 2007; Kerley & Landman 2006). Their wide diet breadth enables them to tolerate foods of low nutritional value and while elephants do not display selective feeding behaviour they are still able to obtain their daily nutritional requirements (Kerley & Landman 2006; Shrader et al. 2012). Elephants prefer feeding on grass as it tends to have a greater palatability, holds a higher nutritional value and is generally lower in tannin concentrations compared to browse species (Guy 1976; Cooper & Owen-Smith 1985; de Boer et al. 2000; Shoshani et al. 2004 Codron et al. 2006). Due to the associated constraints of bulk feeding, elephant’s diets tend toward a frequency dependent strategy, whereby they select habitats and utilise resources or plant species in proportion to their respective abundances within the habitat (Laws 1970b; de Boer et al. 2000; Shrader et al. 2012). As a result, elephants tend to predominantly graze in wetter seasons and predominantly browse in drier seasons (Cerling et al. 1999; de Boer et al. 2000; Cerling et al. 2004; Cerling et al. 2006; Codron et al. 2006; Codron et al. 2011).

There is little consensus regarding the African elephant’s feeding strategy as several authors have argued that elephants diets do not reflect the relative availabilities of browse and grass species, rather elephants strategize their feeding habits to maximise their nutrient uptake and so focus on plants and habitats that have a higher nutritional value (Wing & Buss 1970; Shoshani et al. 2007; Codron et al. 2006; Loarie et al. 2009a; Owen-Smith & Chapota 2012). Studies have demonstrated that when

Stellenbosch University https://scholar.sun.ac.za

feeding, elephants do display preferences for certain tree species and growth forms, such as tree height, and suggest that this could be due to the differences in nutritional content and/or plant defence mechanisms (Barnes 1982; Jachmen & Bell 1985; Stokke & du Toit 2000; Styles & Skinner 2000). There are a limited number of studies that investigate the diet of elephants in relation to food availability and so despite numerous dietary studies, our current understanding of elephant foraging decisions is still theoretical (Jachmen & Bell 1985; O’Conner et al. 2007). This is partly due to the comparison of data from different studies within different habitats and different time periods which can lead to multiple sources of error (Field 1972; Cerling et al. 2009).

Due to the theoretical nature of our understanding of elephant foraging strategies there is still some controversy as to whether elephants are principally browsers or grazers. A few authors have suggested that savanna elephants are predominantly grazers as certain populations had a high component of grass in their diets at the time of study (Laws 1970a; Laws 1970b; Wing & Buss 1970; Cerling et al. 2004). Other authors have found savanna elephants to have a high proportion of browse and very little grass in their diet (Cerling et al. 1999; Cerling et al. 2004; Cerling et al. 2006; Cerling et al. 2009; Codron et al. 2011). Traditional methods of analysing elephant’s diet consisted of field observations or microhistological examination of gut and faecal contents (de Boer et al. 2000). These methods were deemed to be too time consuming to facilitate interregional or interhabitat comparisons of diet and so did little to solve the controversy (de Boer et al. 2000). A newer method that has become a common tool for studying the nutritional ecology of wildlife is stable isotope analysis and is based on knowledge of fractionation occurring in metabolic processes of plants and animals (van der Merwe et al. 1988; Codron et al. 2005). Stable carbon isotope analysis provides a means to quantify the consumption of browse and grass by savanna herbivores across multiple space and time scales (Codron et al. 2006).

Stable carbon isotope ratios are distinct to plants that employ a C3 photosynthetic pathway (trees, shrubs and forbs) and plants that employ a C4 photosynthetic pathway (grasses) (Smith & Epstein 1971; Vogel et al. 1990). The consumer’s tissues and excreta reflect the carbon from the food and so herbivore 13C/12C ratios reflect the proportions of browse and grass in diet (Vogel et al. 1990; Cerling et al. 1999; Codron et al. 2005; Codron et al. 2006). Although stable isotope analysis lacks the finer detail of microhistological techniques, the approach is a standardised method that produces data sets that are easily comparable with other studies across the continent. Stable isotope studies across Africa have shown that elephants are primarily browsers (<15% grass in diets) (van der Merwe et al. 1988; Tiezen et al. 1989; Vogel et al. 1990; Cerling et al. 1999) but moderate amounts of grass are consumed by some populations, such as the Tsavo (hereafter, TENP) and ANP elephant populations isotopically analysed by Koch et al. (1995) and Cerling et al. (1999).

Stellenbosch University https://scholar.sun.ac.za

Although elephants may currently be primarily browsers that substitute their diet with grass, their phylogenetic lineage suggests that modern elephants have ancestors that shifted between grazing and browsing (Cerling et al. 1999). From the late Miocene era to the Pleistocene era elephant’s predominantly grazed but a dietary shift occurred less than one million years ago where elephants switched to browsing (Cerling et al. 1999). Like their ancestors, modern African elephants are large bodied, generalist herbivores that are capable of feeding on both browse (C3 biomass) and grass (C4 biomass) although the proportions of browse: grass consumed will vary according to factors that include geographic variability, rainfall and season, resource availability and quality, gender and proximity to permanent water sources (van der Merwe et al. 1988; Codron et al. 2006; Codron et al. 2011; Shrader et al. 2012).

Geographic Variability Carbon from the vegetation consumed by herbivores is incorporated into animal body tissues and so the stable isotope analysis of elephants bone collagen reflects the 13C/12C ratios of the proportion of grass to browse consumed by an individual (Cerling et al. 1999). The analysis of a number of different bone samples collected from around the Africa revealed that savanna elephants from Namibia, South

Africa, Botswana, Zambia and Malawi have a near pure C3 diet with very little C4 intake (van der Merwe et al. 1988). There were two exceptions to this; the first was AENP, South Africa, where elephants have a relatively low C3 diet due to the CAM plants that make up to 40% of plant cover in the area (van der Merwe et al. 1988; Kerley & Landman 2006). The second was the TENP and ANP populations in Kenya, where elephants had a significantly large proportion of C4 grass in their diet (van der Merwe et al. 1988). In TENP, this was attributed to a high elephant density that transformed the woodlands into grassland (Napier Bax & Sheldrick 1963, van der Merwe et al. 1988; Cerling et al. 2004).

The isotopic analysis of bone collagen tends to under represent the grass component of the diet due to the inefficiency of the elephant’s digestive system at extracting inaccessible proteins from C4 grasses (van der Merwe et al. 1988; Codron et al. 2005; Codron et al. 2006). Despite representing the excreted and undigested portion of the diet, the carbon isotopic composition of faeces has shown to be consistent with dietary values during both experimental and field studies and enables a finer scale of comparison between location and time (Tieszen et al. 1989; Sponheimer et al. 2003; Codron et al. 2005; Codron et al. 2006). The use of faecal carbon in isotopic studies suggests that elephants chiefly browse on C3 woody vegetation in the dry season and graze on C4 grasses in the wet season (Cerling et al. 2004; Cerling et al. 2006; Cerling et al. 2008). A dietary study conducted in the KNP using faecal carbon indicated that elephants inhabiting the northern areas of the park consumed significantly more grass (>30% more) compared to southern elephants, in the dry season (Codron et al. 2006). The difference in grass consumed was a result of the northern elephants consuming more grass as an

Stellenbosch University https://scholar.sun.ac.za

alternative resource to a homogenous mopane-dominated landscape, whereas southern elephant populations had a higher tree diversity and so a greater dietary variety (Codron et al. 2006). The use of faecal samples in isotopic analysis has demonstrated that location does affect the proportion of browse: grass consumed and is linked to climate and resource availability (van der Merwe et al. 1988; Cerling et al. 1999; Shrader et al. 2012).

Seasonal Rainfall and Habitat Selection The quality of the Savanna elephant’s diet fluctuates with the seasonal variation of resources. In the dry season, diet quality decreases for elephant populations due to limited availability and poor quality of resources (de Boer et al. 2000; Woolley et al. 2008; Gaugris et al. 2009). In the wet season, however, the quality and quantity of resources increases and so elephants selectively feed on vegetation that is higher in nutritional value, this constitutes an increase in the consumption of C4 grass by a mixed- feeder that is normally dominated by browse (Barnes 1982; de Boer et al. 2000; Cerling et al. 2004; Cerling et al. 2006; Codron et al. 2006; Cerling et al. 2009; Kos et al. 2012; Shrader et al. 2012). Owen- Smith & Chapota (2012) found that elephants only browsed on 30% of 27 common woody species during the wet season but this increased to 50% of the same species in the dry season. This indicates that elephants prefer high quality forage and so they select woody species based on nutritional quality and not availability (Owen-Smith & Chapota 2012). The elephants also shift the plant parts browsed on from just leaves and shoots in the wet season to stem bark and roots in the dry season (Owen- Smith & Chapota 2012).

Season influences the quality and availability of forage so with the change in season elephants must adapt their foraging strategy (Loarie et al. 2009a; Owen-Smith & Chapota 2012; Shrader et al. 2012). However, elephants select habitats independent of season as they consistently seek out vegetation that is greener than expected throughout the year (Skarpe et al. 2004; Loarie et al. 2009a; Codron et al. 2011; Shrader et al. 2012; Okello et al. 2015). In the wet season elephants select seasonally variable landscapes, which only contain high quality forage for a few months a year such as shrub lands, grasslands and dambos (flooded grasslands) (Loarie et al. 2009a). In the dry season landscapes become less variable as vegetation quality decreases and is interspersed with bare ground and so elephants prefer landscapes of good quality, such as well wooded areas and closed woodlands (Loarie et al. 2009a; Shrader et al. 2012). Thus, elephants initially focus their feeding on large spatial scales, preferring landscapes of higher quality (Loarie et al. 2009a). Yet the dependence of elephant on woody species in the dry season and the narrowed selection of woody species in the wet season suggests that elephants do adjust their foraging strategy on a seasonal basis (Loarie et al. 2009a; Codron et al. 2011; Owen-Smith & Chapota 2012; Shrader et al. 2012).

Stellenbosch University https://scholar.sun.ac.za

Sexual and spatial segregation in habitat use In order to optimise nutritional trade-offs against limitations imposed by body size, nutritional requirements, digestive anatomy, physiology and food resources, mammalian herbivores adopt foraging strategies. Due to their large body size elephants have adopted sexual segregation as a form of foraging strategy (Stokke & du Toit 2000). This is often referred to as the Body Size Hypothesis whereby larger ungulates can tolerate a wider range in diet quality (digestibility) compared to smaller ungulates (Stokke & du Toit 2000). According to the hypothesis, elephant bulls and cows should utilise different browse resources in response to different metabolic demands (Stokke & du Toit 2000; Shannon et al. 2006a; Woolley et al. 2008). Females are expected to feed more selectively on higher quality forage whereas males are expected to feed less selectively on abundant forage of lower quality (Stokke & du Toit 2000; Shannon et al. 2006a; Woolley et al. 2008).

The Body Size Hypothesis proposes that bulls have higher ingestion rates and thus have a greater digestion efficiency than cows (Stokke & du Toit 2000; Shannon et al. 2006b; Woolley et al. 2008). Bulls have a less diverse diet in terms of woody plant species and they consume more plant parts than cows and family units (Stokke & du Toit 2000; Shannon et al. 2006b; Woolley et al. 2008). Additionally, bulls spend more time foraging on each woody plant, ingesting greater quantities of forage at each feeding bout and displaying destructive behaviour by engaging in tree felling (Guy 1976; Barnes 1982; Stokke & du Toit 2000; Shannon et al. 2006b; Woolley et al. 2008). Smaller bodied cows, specifically breeding individuals, have higher mass specific nutritional demands due to gestation and lactation and so feed more selectively on higher quality forage than bulls (Stokke & du Toit 2000; Shannon et al. 2006b; Woolley et al. 2008). In general, cows tend to have a more diverse diet in terms of woody plant species, are more selective for higher quality plant parts and exhibit shorter feeding bouts due to the social constraints of family units compared to bulls (Guy 1976; Barnes 1982; Stokke & du Toit 2000; Shannon et al. 2006b; Woolley et al. 2008).

The ranging behaviour of bulls and family units differs depending on the season. In a study focusing on the sexual segregation in habitat use of elephant in Botswana, Stokke and du Toit (2002) reported in the dry season bulls frequented more habitat types compared to family units and roamed widely (>10km) from perennial drinking water (Stokke & du Toit 2002). Larger bodied male elephants are consistently able to adapt their movement behaviour according to forage quality and abundance (Ntumi et al. 2005; Shannon et al. 2006b; de Knegt et al. 2011), whereas the smaller bodied females are unable to exhibit the same flexibility in ranging behaviour because of their higher nutritional demands, lower tolerance to fibrous forage and the social and energetic constraints of group living with juveniles and calves (Stokke & du Toit 2000; Shannon et al. 2006b; Woolley et al. 2008; de Knegt et al. 2011). Due to body size constraints and group living, elephants adopt a selective strategy that

Stellenbosch University https://scholar.sun.ac.za

maximises the rate of nutritional intake by targeting abundant but often lower quality foraging opportunities through the expansion of home ranges during years when rainfall is below average (Shannon et al. 2006b; Kinahan et al. 2007; Smit et al. 2007; de Knegt et al. 2011). This indicates that home range size is primarily dictated by abiotic factors at a landscape scale, more specifically though home range size is influenced by vegetation productivity only during the wet season as elephants are less restricted by water compared to the dry season (Leggett 2006; Shannon et al. 2006b; Chamaille- Jammes et al. 2007; Harris et al. 2008; Roux & Bernard 2007; Young et al. 2009a).

Habitat Use and proximity to water resources African elephant movement is complex and seasonally variable but in general habitat use tends to increase with proximity to water sources (Harris et al. 2008; Staub et al. 2013). The distance travelled by elephants within seasons varies greatly depending on the resources available (Grainger et al. 2005; Chamaille-Jammes et al. 2007). When essential resources occur locally elephants will travel less than a few kilometres from day to day (Chamaille-Jammes et al. 2007; Harris et al. 2008). Elephant ranging behaviour also depends on the season: in the dry season elephants are limited to areas surrounding permanent water sources whereas in the wet season ephemeral pools allow elephants to expand their movement across more of the landscape (Bhima & Dudley 1997; Dudley et al. 2001; Galanti et al. 2006; Leggett 2006; Chamaille-Jammes et al. 2007; Kinahan et al. 2007; de Beer & van Aarde 2008; Harris et al. 2008; Loarie et al. 2009a; Young et al. 2009b). However, this is not always the case as in Etosha National Park for example, the vegetation is much sparser, low in productivity and species poor, so elephants must travel greater distances between vegetation and water sources (Verliden & Gavor 1998; de Beer et al. 2006; Harris et al. 2008). A similar example of this is in Pongola Reserve, South Africa, where elephants displayed smaller ranges in the wet season compared to the dry season due to abundant vegetation that proved it unnecessary for elephants to increase the size of their home range (Shannon et al. 2010). This indicates that elephant home ranges are dependent upon the season, the vegetation available and the distance to water sources (Whitehouse & Schoeman 2002; Leggett 2006; Chamaille-Jammes et al. 2007; Roux & Bernard 2007; Harris et al. 2008).

Conclusion

Across Africa elephant numbers are decreasing due to habitat loss and fragmentation, human- elephant conflict, poaching for meat and ivory and the negative localised impacts of elephants on their habitat (Blanc et al. 2002; Blanc et al. 2007). In Southern Africa, a growing number of reserves are being fenced, essentially creating ecological islands in the form of artificially closed systems (Boone & Thompson Hobbs 2004). Fences have multiple effects on ecological processes within protected areas including limiting wildlife dispersal and the resources available to wildlife populations (Newmark

Stellenbosch University https://scholar.sun.ac.za

2008). Many elephant populations are increasing in these fenced, protected areas due to increased protection, limited migration and the provision of resources in the form of artificial waterholes (Laws 1970b; van Aarde & Jackson 2007). This complicates the management of these structured systems as the vegetation is exposed to increased foraging pressure that potentially leads to loss of woody vegetation cover (Duffy et al. 2002; van Aarde & Jackson 2007; Loarie et al. 2009a; Mapaure & Moe 2009). In order to maintain the balance between healthy elephant populations and the integrity of the ecosystems they occupy it has become a necessity for reserve managers to understand the primary factors affecting the demographic vital rates of elephant populations and how these can be manipulated to control population growth (Moss 2001; Wittemyer et al. 2005; Gough & Kerley 2006; Wittemyer et al. 2014). The vital rates of populations are primarily affected by factors such as resource quality and availability and so it is important to investigate the species dependence on the resources and their basic nutritional requirements, such as diet, in order to compile an effective management plan.

Stellenbosch University https://scholar.sun.ac.za

References:

 Barnes, R.F.W. (1982) Elephant feeding behaviour in Rusha National Park, Tanzania. African Journal of Ecology, 20, 123-136.  Barnes, R.F.W. (1983) The Elephant Problem in Rusha National Park, Tanzania. Biological Conservation, 26, 127-148.  Bhima, R. & Dudley, C.O. (1997) The influence of the Shire River on Liwonde National Park, Malawi, with special reference to elephant movements. Koedoe, 40(2), 9-18.  Blake, S. & Hedges, S. (2004) Sinking the Flagship: the Case of Forest Elephants in Asia and Africa. Conservation Biology, 18(5), 1191-1202.  Blanc, J.J., Barnes, R.F.W., Craig, G.C., Dublin, H.T., Thouless, C.R., Douglas-Hamilton, I. & Hart, J.A. (2007) African Elephant Status Report 2007: and update from the African Elephant Database. Occasional Paper Series of the IUCN Species Survival Commission, No 33. IUCN/SSC African Elephant Specialist Group IUCN, Gland, Switzerland. Vi +276 pp.  Blanc, J.J., Thouless, C.R., Hart, J.A., Dublin, H.T., Douglas-Hamilton, I., Craig, C.G. & Barnes, R.F.W. (2002) African Elephant Status Report 2002: An update from the African Elephant Database. IUCN/SSC African Elephant Specialist Group. IUCN. Gland, Switzerland and Cambridge, UK. Vi + 304pp.  Boone, R.B. & Hobbs, T. (2004) Lines around fragments: effects of fencing on large herbivores. African Journal of Range and Forage Science, 21(3), 147 -156.  Brown, J.H. (1984) On the Relationship between Abundance and Distribution of Species. The American Naturalist, 124, 255-279.  Caughley, G. (1976) The elephant problem – an alternative hypothesis. East African Wildlife Journal, 14, 265-283.  Cerling, T.E., Harris, J.M. & Leakey, M.G. (1999) Browsing and Grazing in Elephants: The Isotope record of Modern and Fossil Proboscideans. Oecologia, 120(3), 364–374.  Cerling, T.E., Passey, B.H., Ayliffe, L.K., Cook, C.S., Ehleringer, J.R., Harris, J.M., Dhidha, M.B. & Kasiki, S.M. (2004) Orphans’ tales: seasonal dietary changes in elephants from Tsavo National Park, Kenya. Palaeoecology, 206, 367 – 376.  Cerling, T.E., Wittemyer, G., Ehleringer, J.R., Remlen, C.H. & Douglas-Hamilton, I. (2009) History of animals using Isotopes Records (HAIR): A 6-year dietary history of one family of African Elephants. PNAS, 106(20), 8093 – 8100.

Stellenbosch University https://scholar.sun.ac.za

 Cerling, T.E., Wittemyer, G., Rasmussen, H.B., Vollrath, F., Cerling, C., Robinson, T.J. & Douglas- Hamilton, I. (2006) Stable isotopes in elephant hair document migration patterns and diet changes. PNAS, 103(2), 371-373.  Chamaille-Jammes, S., Fritz, H., Valeix, M., Murindagomo, F. & Clobert, J. (2008) Resource variability, aggregation and direct density dependence in an open context: the local regulation of an African elephant population. Journal of Animal Ecology, 77, 135-144.  Chamaille-Jammes, S., Valeix, M. & Fritz, H. (2007) Managing heterogeneity in elephant distribution: interactions between elephant population density and surface water availability. Journal of Applied Ecology, 44, 625-633.  Clauss, M., Streich, W.J., Schwarm, A., Ortmann, S. & Hummel, J. (2007) The relationship of food intake and ingesta passage predicts feeding ecology in two different megaherbivore groups. Oikos, 116, 209-216.  Clutton-Brock, T. (2002) Breeding together: kin selection and mutualism in cooperative vertebrates. Science, 269, 69-72.  Clutton-Brock, T.H., Major, M., Albon, S.D. & Guinness, F.E. (1987) Early Development and Population Dynamics in Red Deer. I. Density-Dependent Effects on Juvenile Survival. Journal of Animal Ecology, 56, 53-67.  Codron, D., Codron, J., Sponheimer, M., Lee-Thorp, J.A., Robinson, T., Grant, C.C. & de Ruiter, D. (2005) Assessing diet in savanna herbivores using stable carbon isotope ratios of faeces. Koedoe, 48, 115-124.  Codron, J., Codron, D., Lee-Thorp, J.A., Sponheimer, M., Kirkman, K., Duffy, K.J. & Sealy, J. (2011) Landscape-scale feeding patterns of African elephants inferred from carbon isotope analysis of faeces. Oecologia, 165, 89-99.  Codron, J., Lee-Thorp, J.A., Sponheimer, M., Codron, D., Grant, R.C. & De Ruiter, D.J. (2006) Elephant (Loxodonta africana) Diets in Kruger National Park, South Africa: Spatial and Landscape Differences. Journal of Mammalogy, 87(1), 27-34.  Cooper, S.M. & Owen-Smith, N. (1985) Condensed Tannins Deter Feeding by Browsing Ruminants in a South African Savanna. Oecologia, 67, 142-146.  Cumming, D.H.M., Fenton, M.B., Rautenbach, I.L., Taylor, R.D., Cumming, G.S., Cumming, M.S., Dunlop, J.M., Ford, G.S., Hovorka, M.D., Johnston, D.S., Kalcounis, M.C., Mahlanga, Z., and C.V. Portfors. (1997) Elephants, woodlands and biodiversity in miombo woodland in Southern Africa. South African Journal of Science, 93, 231-236.

Stellenbosch University https://scholar.sun.ac.za

 de Beer, Y. & van Aarde, R.J. (2008) Do landscape heterogeneity and water distribution explain aspects of elephant home range in Southern Africa’s arid savannas?. Journal of Arid Environments, 72, 2017-2025.  de Beer, Y. Kilian, W., Versfeld, W. & van Aarde, R.J. (2006) Elephants and low rainfall alter woody vegetation in Etosha National Park, Namibia. Journal of Arid Environment, 64, 412 – 421.  de Boer, W.F., Ntumi, C.P., Correia, A.U. & Mafuca, J.M. (2000) Diet and distribution of elephant in the Maputo Elephant Reserve, Mozambique. African Journal of Ecology, 38, 188- 201.  de Boer, W.F., van Langevelde, F., Prins, H.H.T., de Ruiter, C., Blanc, J., Vis, M.J.P., Gaston, K.J. & Douglas-Hamilton, I. (2013) Understanding spatial differences in African elephant densities and occurrence, a continent-wide analysis. Biological Conservation, 159, 468-476.  de Knegt, H.J., van Langevelde, F., Skidmore, A.K., Delsink, A., Slotow, R., Henley, S., Bucini, G., de Boer, W.F., Coughenour, M.B., Grant, C.C., Heitkonig, I.M.A., Henley, M., Knox, N.M., Kohi, E.M., Mwakiwa, E., Page, B.R., Peel, M., Pretorius, Y., van Wieren, S.E. & Prins, H.H.T. (2011) The spatial scaling of habitat selection by African elephants. Journal of Animal Ecology, 80, 270-281.  de Kroon, H., van Groenendael, J. & Ehrlen, J. (2000) A review of methods and model limitations. Ecology, 81, 607-618.  Dudley, J.P., Craig, G.C., Gibson, D.S., Haynes, G. & Kilmowicz, J. (2001) Drought mortality of bush elephants in Hwange National Park, Zimbabwe. African Journal of Ecology, 39, 187-195.  Duffy, K.J., van Os, R., Vos, S., van Aarde, J. Ellish, G. & Stretch, A-M.B. (2002) Estimating impact of reintroduced elephant on trees in a small reserve. South African Journal of Wildlife Research, 32(1), 23-29.  Eberhardt, L.L. (2002) A paradigm for population analysis of long-lived vertebrates. Ecology, 83, 2841-2852.  Field, C.R. (1972) The food habits of wild ungulates in Uganda by analyses of stomach contents. African Journal of Ecology, 10, 17-42.  Freeman, E.W., Whyte, I. & Brown, J.L. (2008) Reproductive evaluation of elephants culled in Kruger National Park, South Africa between 1975 and 1995. African Journal of Ecology, 47, 192-201.  Gaillard, J., Festa-Bianchet, M. & Yoccoz, N.G. (1998) Population dynamics of large herbivores: variable recruitment with constant adult survival. Trends in Ecology and Evolution, 13, 58-62.

Stellenbosch University https://scholar.sun.ac.za

 Gaillard, J.M., Festa-Bianchet, M., Delorme, D. & Jorgenson, J. (2000) Body mass and individual fitness in female ungulates: bigger is not always better. Proceedings of the Royal Society of London B, 267, 471-477.  Galanti, V., Preatoni, D., Martinoli, A., Wauters, L.A. & Tosi, G. (2006) Space and habitat use of the African elephant in the Tarangire-Manyara ecosystem, Tanzania: Implications for conservation. Mammalian Biology, 71(2), 99-114.  Gaugris, J.Y. & van Rooyen, M.W. (2009) Effects of water dependence on the utilisation pattern of woody vegetation by elephants in the Tembe Elephant Park, Maputaland, South Africa. African Journal of Ecology, 48, 126-134.  Gillson, L. & Lindsay, K. (2003) Ivory and ecology – changing perspectives on elephant management and the international trade in ivory. Environmental Science, 6, 411-419.  Gough, K.F. & Kerley, G.I.H. (2006) Demography and population dynamics in the elephants Loxodonta africana of Addo Elephant National Park, South Africa: is there evidence of density dependent regulation? Oryx, 40(4), 434-441.  Grainger, M., van Aarde, R. & Whyte, I. (2005) Landscape heterogeneity and the use of space by elephants in the Kruger National Park, South Africa. African Journal of Ecology, 43, 369- 375.  Guldemond, R. & van Aarde, R. (2007) The impact of elephants on plants and their community variables in South Africa’s Maputaland. African Journal of Ecology, 45, 327-335.  Guldemond, R. & Van Aarde, R. (2008) A Meta-Analysis of the Impact of African Elephants on Savanna Vegetation. Journal of Wildlife Management, 72(4), 892-899.  Guldemond, R.A.R., Lehman, E.R., Ferreira, S.M. & van Aarde, R.J. (2005) Elephant numbers in Kafue National Park, Zambia. Pachyderm, 39, 50-54.  Guy, P.R. (1976). The feeding behaviour of elephant (Loxodonta africana) in the Sengwa Area, Rhodesia. South African Journal of Wildlife Research, 6(1), 55-63.  Harris, G.M., Russeli, G.J., van Aarde, R.I. & Pimm, S.L. (2008) Rules of habitat use by elephants Loxodonta africana in Southern Africa: insights for regional management. Oryx, 42 (1), 66-75.  Hayward, M.W. & Kerley, G.I.H. (2009) Fencing for conservation: Restriction of evolutionary potential or a riposte to threatening processes? Biological Conservation, 142, 1-13.  Hoare, R.E. & du Toit, J.T. (1999) Coexistence between People and Elephants in African Savannas. Conservation Biology, 13(3), 633-639.  Jacobs, O.S. & Biggs, R. (2002) The impact of the African elephant on marula trees in the Kruger National Park. South African Journal of Wildlife Research, 32(1), 13-22.

Stellenbosch University https://scholar.sun.ac.za

 Jones, M. (2016) More Highs than Lows for Wildlife at CITES Meeting. National Geographic. Washington, D.C. United States of America. Available from: http://voices.nationalgeographic.com/2016/10/06/more-highs-than-lows-for-wildlife-at- cites-meeting/ (Accessed October 2016).  Kerley, G.I.H. & Landman, M. (2006) The impacts of elephants on biodiversity in the Eastern Cape Subtropical Thickets. South African Journal of Science, 102, 395-402.  Kinahan, A.A., Pimm, S.L. & van Aarde, R.J. (2007) Ambient temperature as a determinant of landscape use in savanna elephant, Loxodonta africana. Journal of Thermal Biology, 32, 47- 58.  Kioko, J., Zink, E., Sawdy, M. & Kiffner, C. (2013) Elephant (Loxodonta africana) demography and behaviour in the Tarangire-Manyara Ecosystem, Tanzania. South African Journal of Wildlife Research, 43(1), 44-51.  Koch, P.L., Heisinger, J., Moss, C., Carlson, R.W., Fogel, M.L. & Behrensmeyer. (1995) Isotopic Tracking of Change in Diet and Habitat Use in African Elephants. Science, 267, 1340-1343.  Kos, M., Hoetmer, A.J., Pretorius, Y., de Boer, W.F., de Knegt, H., Grant, C.C., Kohi, E., Page, B., Peel, M., Slotow, R., van der Waal, C., van Wieren, S.E., Prins, H.H.T. & van Langevelde, F. (2012) Seasonal diet changes in elephant and impala in mopane woodland. European Journal of Wildlife Research, 58, 279-287.  Laws, R.M. (1970a) Elephants and habitats in north Bunyoro, Uganda. African Journal of Ecology, 8, 163-180.  Laws, R.M. (1970b) Elephants as agents of habitat and landscape change in East Africa. Oikos, 21, 1-15.  Leggett, K.E.A. (2006) Home range and seasonal movement of elephants in the Kunene Region, north-western Namibia. African Zoology, 41(1), 17-36.  Lewis, D.M. (1984) Demographic changes in the Luangwa valley elephants. Biological Conservation, 29, 7-14.  Loarie, S.R., van Aarde, R.J. & Pimm, S.L. (2009a) Elephant seasonal vegetation preferences across dry and wet savannas. Biological Conservation, 142(12), 3099-3107.  Loarie, S.R., van Aarde, R.J. & Pimm, S.L. (2009b) Fences and artificial water affect African savannah elephant movement patterns. Biological Conservation, 142(12), 3086-3098.  Lombard, A.T., Johnson, C.F., Cowling, R.M. & Pressey, R.L. (2001) Protecting plants from elephants: botanical reserve scenarios within the Addo Elephant National Park, South Africa. Biological Conservation, 102, 191-203.

Stellenbosch University https://scholar.sun.ac.za

 Lovett, J.C. (2009) Elephants and the conservation dilemma. African Journal of Ecology, 47, 129-130.  Mapaure, I. & Moe, S.R. (2009) Changes in the structure and composition of miombo woodlands mediated by elephants (Loxodonta africana) and fire over a 26-year period in north-western Zimbabwe. African Journal Ecology, 47, 175-183.  Martin, J., Chamaille-Jammes, S., Nichols, J.D, Fritz, H., Hines, J.E, Fonnesbeck,C.J., Mackenzie, D.I. & Bailey, L. (2010) Simultaneous modelling of habitat suitability, occupancy, and relative abundance: African elephants in Zimbabwe. Ecological Applications, 20, 1173-1182.  Moss, C.J. (2001) The demography of an African elephant (Loxodonta africana) population in Amboseli, Kenya. Journal of Zoology, 255, 145-156.  Murdoch, W.W. (1994) Population Regulation in Theory and Practise. Ecology, 75, 271-287.  Napier Bax, P. & Sheldrick, D.L.W. (1963) Some preliminary observations on the food of elephant in the Tsavo Royal National Park (East) of Kenya. African Journal of Ecology, 1(1), 40- 51.  Neubert, M.G. & Caswell, H. (2000) Density-dependent vital rates and their population dynamic consequences. Journal of Mathematical Biology, 41, 103-121.  Newmark, W.D. (2008) Isolation of African Protected Areas. Frontiers in Ecology and the Environment, 6(6), 321-328.  Ntumi, C.P., van Aarde, R.J., Fairall, N. & de Boer, W.F. (2005) Use of space and habitat by elephants (Loxodonta africana) in the Maputo Elephant Reserve, Mozambique. South African Journal of Wildlife Research, 35(2), 139-146.  O’Connor, T., Goodman, P.S. & Clegg, B. (2007) A functional hypothesis of the threat of local extirpation of woody plant species by elephant in Africa. Biological Conservation, 136, 329- 345.  Okello, M.M., Njumbi, S.J., Kiringe, J.W. & Isiiche, J. (2015) Habitat use and preference by the African elephant outside of protected area, and management implications in the Amboseli Landscape, Kenya. International Journal of Biodiversity and Conservation, 7(3), 211-236.  Osborn, F.V. & Parker, G.E. (2003) Towards an integrated approach for reducing the conflict between elephants and people: a review of current research. Oryx, 37(1), 80-84.  Ostfeld, R.S. & Keesing, F. (2000) Pulsed resources and community dynamics of consumers in terrestrial ecosystems. Trends in Ecology and Evolution, 15, 232-237.  Owen-Smith, N. & Chapota, J. (2012) Selective feeding by a megaherbivore, the African elephant (Loxodonta africana). Journal of Mammalogy, 93(3), 698-705.

Stellenbosch University https://scholar.sun.ac.za

 Roux, C. & Bernard, R.T.F. (2007) Home range size, spatial distribution and habitat use of elephants in two enclosed game reserves in the Eastern Cape Province, South Africa. African Journal of Ecology, 47, 146-153.  Shannon, G., Page, B., Duffy, K.J. & Slotow, R. (2010) The ranging behaviour of a large sexually dimorphic herbivore in response to seasonal and annual environmental variation. Austral Ecology, 35, 731-742.  Shannon, G., Page, B., Slotow, R. & Duffy, K. (2006a) African elephant home range and habitat selection in Pongola Game Reserve, South Africa. African Zoology, 41 (1), 37-44.  Shannon, G., Page, B.R., Duffy, K.J. & Slotow, R. (2006b) The role of foraging behaviour in the sexual segregation of the African elephant. Oecologia, 150, 344-354.  Shoshani, J., Hagos, Y., Yacob, Y., Ghebrehiwet, M. & Kebrom, E. (2004) Elephants (Loxodonta africana) of Zoba Gash Barka, Eritrea: Part 2. Numbers and distribution, ecology and behaviour and, fauna and flora in their ecosystem. Pachyderm, 36, 52-68.  Shrader, A.M., Bell, C., Bertolli, L. & Ward, D. (2012) Forest of the trees: At what scale do elephants make foraging decisions? Acta Oecologica, 42, 3-10.  Shrader, A.M., Pimm, S.L, & van Aarde, R.J. (2010). Elephant survival, rainfall and the confounding effects of water provision and fences. Biodiversity Conservation, 19, 2235-2245.  Sitati, N.W., Walpole, M.J. & Leader-Williams, N. (2005) Factors affecting the susceptibility of farms to crop raiding by African Elephants: using a predictive model to mitigate conflict. Journal of Applied Ecology, 42, 1175-1182.  Skarpe, C., Aarrestad, P.A., Andreassen, H.P., Dhillion, S.S., Dimakatso, T., du Toit, J.T., Halley, D.J., Hytteborn, H., Makhabu, S., Mari, M., Marokane, W., Masunga, G., Modise, D., Moe, S.R., Mojaphoko, R., Mosugelo, D., Motsumi, M., Neo-Mahupeleng, G., Ramotadima, M., Rutina, L., Sechele, L. & Sejoe, T.B., Stokke, S., Swenson, J.E., Taolo, C., Vandewalle, M. & Wegge, P. (2004) The return of the giants: Ecological Effects of an Increasing Elephant Population. Ambio, 33(6), 276-282.  Slotow, R., Garai, M.E., Reilly, B., Page, B.& Carr, R.D. (2005) Population dynamics of elephants re-introduced to small fenced reserves in South Africa. South African Journal of Wildlife Research, 35(1), 23-32.  Smit, I.P.J, Grant, C.C. & Whyte, I.J. (2007) Elephants and water provision: what are the management links? Diversity and Distributions, 13, 666-669.  Smith, B. & Epstein, S. (1971) Two Categories of 13C/12C Ratios for Higher Plants. Plant Physiology, 47, 380-384.

Stellenbosch University https://scholar.sun.ac.za

 Sponheimer, M., Robinson, T., Ayliffe, L., Passey, B., Roeder, B., Shipley, L., Lopez, E., Cerling, T., Dearing, D. & Ehleringer, J. (2003) An experimental study of carbon-isotope fractionation between diet, hair, and faeces of mammalian herbivores. Canadian Journal of Zoology, 81, 871-876.  Staub, C.G., Binford, M.W. & Stevens F.R. (2013) Elephant herbivory in Majete Wildlife Reserve, Malawi. African Journal of Ecology, 51(4), 536-543.  Stiles, D. (2004) The ivory trade and elephant conservation. Environmental Conservation, 31(4), 309-321.  Stokke, S. & du Toit, J.T. (2000) Sex and size related differences in the dry season feeding patterns of elephants in Chobe National Park, Botswana. Ecography, 23, 70-80.  Stokke, S. & du Toit, J.T. (2002) Sexual segregation in habitat use by elephants in Chobe National Park, Botswana. African Journal of Ecology, 40, 360-371.  Styles, C.V. & Skinner, J.D. (2000) The influence of large mammalian herbivores on growth form and utilization of mopane trees, Colophospermum mopane, in Botswana’s Northern Tuli Game Reserve. African Journal of Ecology, 38, 95-101.  Tieszen, L.L., Boutton, T.W., Ottichilo, W.K., Nelson, D.E. & Brandt, D.H. (1989) An assessment of long-term food habits of Tsavo elephants based on stable carbon and nitrogen isotope rations of bone collagen. African Journal of Ecology, 27, 219-226.  Trimble, M.J., Ferreira, S.M., van Aarde, R.J. (2009) Drivers of megaherbivores demographic fluctuations: inference from elephants. Journal of Zoology, 279, 18-26.  Tuljapurkar, S., Horvitz, C.C. & Pascarella, J.B. (2003) The many growth rates and elasticities of populations in random environments. The American Naturalist, 162, 489-502.  van Aarde, R.J. & Jackson, T.P. (2007) Megaparks for metapopulations: Addressing the causes of locally high elephant numbers in Southern Africa. Biological Conservation, 134, 289-297.  van Aarde, R.J., Whyte, I. & Pimm, S. (1999) Culling and the dynamics of the Kruger National Park African elephant population. Animal Conservation, 2, 287-294.  van der Merwe, N.J., Lee Thorp, J.A. & Bell, R.H.V. (1988) Carbon isotopes as indicators of elephant diets and African environments. African Journal of Ecology, 26, 163-172.  Vogel, J.C., Eglington, B. & Auret, J.M. (1990) Isotope fingerprints in elephant bone and ivory. Nature, 346, 747-748.  Whitehouse, A.M. & Hall-Martin, A.J. (2000) Elephants in Addo Elephant National Park, South Africa: reconstruction of the population’s history, Oryx, 34(1), 46-55.  Whitehouse, A.M. & Schoeman, D.S. (2002) Ranging behaviour of elephants within a small fenced area in Addo Elephant National Park, South Africa. African Zoology, 38(1), 95-108.

Stellenbosch University https://scholar.sun.ac.za

 Whyte, I., van Aarde, R. & Pimm, S.L. (1998) Managing the elephants of Kruger National Park. Animal Conservation, 1, 77-83.  Whyte, I.J., van Aarde, R.J. and Pimm, S.L. (2003) Kruger’s Elephant Population: Its size and Consequences for Ecosystem Heterogeneity. The Kruger Experience. (eds J.T. du Toit, K.H. Rogers & H.C. Biggs), pp: 332-348. Island Press. Washington.  Wing, L.D. & Buss, I.O. (1970) Elephants and Forests. Wildlife Monographs, 19, 3-92.  Wittemyer, G., Barner Rasmussen, H. & Douglas-Hamilton, I. (2007a) Breeding phenology in relation to NDVI variability in free-ranging African elephant. Ecography, 30, 42-50.  Wittemyer, G., Daballen, D. and Douglas-Hamilton, I. (2013) Comparative Demography of an At-Risk African Elephant Population. Plos One, 8(1), e53726.  Wittemyer, G., Daballen, D., Rasmussen, H., Kahindi, O. & Douglas-Hamilton, I. (2005) Demographic status of elephants in the Samburu and Buffalo Springs National Reserves, Kenya. African Journal of Ecology, 43, 44-47.  Wittemyer, G., Granswindt, A. & Hodges, K. (2007b) The impact of ecological variability on the reproductive endocrinology of wild female African elephants. Hormones and Behaviour, 51, 346-354.  Woolley, L-A., Millspaugh, J.J., Woods, R.J., van Rensburg, S.J., Page, B.R. & Slotow, R. (2008) Intraspecific Strategic Responses of African Elephants to Temporal Variation in Forage Quality. Journal of Wildlife Management, 73(6), 827-835.  Yang, L.H., Bastow, J.L., Spence, K.O. & Wright, A.N. (2008) What can we learn from resource pulses? Ecology, 89, 621-634.  Young, K.D., Ferreira, S.M. & van Aarde, R.J. (2009a) Elephant spatial use in wet and dry savannas of Southern Africa. Journal of Zoology, 278(3), 189-205.  Young, K.D., Ferreira, S.M. & van Aarde, R.J. (2009b) The influence of increasing population size and vegetation productivity on elephant distribution in the Kruger National Park. Austral Ecology, 34, 329-342.

Stellenbosch University https://scholar.sun.ac.za

Chapter 3

The population structure and habitat use of the African elephant (Loxodonta africana) in Majete Wildlife reserve, Malawi

F.A. Forrer 1, A. Leslie1

1Department of Conservation Ecology and Entomology, Stellenbosch University, Matieland, Western Cape, 7602, South Africa

Abstract

African elephants are a species of high conservation concern. Due to elephant’s status as a keystone species, their economic value and the threat of habitat loss they are the focus of numerous studies across various disciplines. The aim of this study was to firstly determine, with the use of individual identification techniques, the size, sex and age structure of the elephant population in Majete Wildlife Reserve, Malawi. Secondly, this study aimed to determine habitat use and movement of the population over time. An aerial survey conducted in late 2015 counted a total of 389 elephants in the reserve. The present study was able to formally identify 366 of these individuals, consisting of 47 different family herds, 91 solitary bulls and a sex ratio of 1:1. The largest age classes in the population were the medium sized adults (26.8%) and calves (23.8%). In the absence of migration and heavy poaching, the population has an estimated growth rate of 13.8%. In the dry seasons elephants appeared to remain close to perennial water sources such as the rivers, waterholes and springs situated on the eastern side of the reserve. Elephants tended to utilise the water points more in the dry season compared to the wet season and preferred areas of lower altitude mixed woodland. The population has an extremely high growth rate and it is postulated that it is operating under optimal conditions and has not yet begun to stabilise. Due to an increasing density of elephants, their home ranges are now expanding into the least preferred vegetation types, such as Brachystegia dominated woodland, and higher altitude regions in the reserve.

Key words: African elephant, camera traps, demography, habitat use, Malawi

Stellenbosch University https://scholar.sun.ac.za

Introduction

Detailed demographic data is only available for a small number of species, which are typically of high economic or conservation value (Gaillard et al. 1998). Furthermore, demographic data that has been compiled for these species has been generated from a single population and so data regarding multiple populations is generally not available (Gaillard et al. 2000). This essentially limits the ability to monitor a species’ demographic responses to anthropogenic and ecological pressures as there is a lack of information on life history strategies across biological systems (Stockwell et al. 2003; Owen- Smith et al. 2005). Therefore, in order to manage and conserve species in the future it is of the utmost importance that species specific responses are understood, particularly for those species that are threatened or are economically prominent to their range states.

One of the better studied species in terms of detailed demographic data is the African elephant (Loxodonta africana). To date multiple studies have been conducted using individual based monitoring programs or culling data to provide demographic parameters for various populations (Wittemyer et al. 2014). African elephants are a keystone species and the state of their populations are critical to the ecosystems they occupy (Power et al. 1996). The species plays a pivotal role in ecological systems by influencing factors such as canopy cover (Dublin et al. 1990), species diversity and distributions (Pringle 2008) and seed dispersal (Spanbauer & Adler 2015). Although the majority of studies conducted suggest that elephants can have a detrimental effect on their environment as they are capable of transforming habitats, elephants can also be largely beneficial (Gough & Kerley 2006; Morrison et al. 2016). The impact this species has on its environment is determined by a number of factors that include location, available resources, the confinement of a population and population density (Skarpe et al. 2004; Kerley & Landman 2006; Chamaille-Jammes et al. 2008; Loarie et al. 2009). In order to understand what particular factors contribute to the influence an elephant population has on its environment, the demographic responses of the population need to be determined and these data are invaluable for both theoretical and management applications.

The African elephant is a flagship species that is of high economic value, especially in the commercial trade for ivory and ecotourism (Douglas-Hamilton 1987; Owen-Smith et al. 2006; Wittemyer et al. 2014). The species conservation status varies across the African continent, for example nearly a quarter of the total population of elephants inhabits one region in Botswana whereas the species has gone locally extinct in Central and Western Africa (Blanc et al. 2007; Bouche et al. 2011). Concurrently, the African elephant has an extreme life history strategy with long periods of sexual reproduction, sexual immaturity, the longest mammalian gestation period and a very slow rate of reproduction (Moss 2001; Wittemyer et al. 2014). Thus, the study of the population dynamics of such a species is

Stellenbosch University https://scholar.sun.ac.za

limited by the collection of longitudinal data on demography as it would take up to 60 years to gain the complete life history of a single population (Moss 2001). However, due to the high economic value and conservation significance of the African elephant this limitation has not prevented multiple studies of the species being undertaken over the years.

A few of the earliest demographic studies of African mammals focused on elephants and the majority of these studies were conducted in Eastern and Southern Africa (Buss & Smith 1966, Laws 1966, Short 1966, Hanks 1969, Hanks 1972, Williamson 1976). Most of these earlier studies employed the method of carcass examination whereby the carcasses of individuals killed from culling operations were analysed, representing a cross section of a population from a single point in time (Whyte & Grobler 1998; Moss 2001; Whyte et al. 2003). In reserves that employed regular culling as a standard measure to control elephant populations, studies were able to model the demographic data of populations at successive points in time (Whyte & Grobler 1998; Moss 2001; Whyte et al. 2003). However, the current methods used to estimate the size of elephant populations are less invasive and include aerial surveys, faecal counts, direct counts using line transect sampling or stratification, and intensive ground based surveys providing total counts by means of registration of individual known animals (Douglas- Hamilton 1996, Mbugva 1996, Jackman 1996, Barns 1996, Whitehouse & Hall-Martin 2001). Techniques used to determine the age structure of populations involve ground and aerial surveys which rely on methods such as the measurement of shoulder height, back length, foot print length and visual assessments based on visual estimates (Croze 1972, Leuthold 1976, Jachman & Bell 1984, Lindeque 1991). Culling has been eliminated as a standard form of management in many Southern African reserves since the mid 1990’s and so current methods used to gauge the demographic status of a population rely on assumptions and predictions to describe a population’s history and project its future trends.

To date very few studies have followed elephant populations in detail over time. One particular study conducted in Addo Elephant National Park (AENP) recorded the precise population numbers and reconstructed each individual’s life history since the park was fenced in 1931 (Whitehouse & Hall- Martin 2000). Apart from a brief population study conducted in the Luangwa Valley, Zambia, (Lewis 1984) the Amboseli National Park (hereafter, ANP) elephant study in Kenya remains the only other long term study of individually known free ranging elephants (Moss 2001). However, several other studies have employed the same mark-recapture methods used in these long term studies as a management tool to determine the demographic structure of confined elephant populations (Whitehouse & Hall-Martin 2000; Morley & van Aarde 2007; Trimble et al. 2009).

Stellenbosch University https://scholar.sun.ac.za

In Majete Wildlife Reserve (MWR hereafter), Malawi, a total of 213 individual elephants were reintroduced into the park by 2010. Since the reintroduction, aerial counts have suggested an increasing elephant population but no detailed demographic data has ever been recorded. Thus, reserve management is lacking the vital information it needs to decide a future course of action for this population.

The primary aim of this study was to characterise the demographic status of the Majete elephant population. More specifically the study aimed to firstly, use individual identification techniques to determine population size, age and sex structure. Secondly, to compare the populations current demographic parameters with those from initial reintroductions. Thirdly, this study aimed to determine the habitat use and movement of the elephants throughout MWR.

Techniques used to estimate the population size or densities of large mammals living in wooded areas are poorly developed and methods such as faecal counts and aerial surveys are inadequate for a reserve, such as MWR, where the majority of the vegetation is woodland (Caro 1999; Walsh et al. 2001; Whitehouse et al. 2001; Payne et al. 2003). Although aerial survey methods are well developed for open habitats, they tend to underestimate the number of individuals in small populations, below 250 individuals, in dense habitats and this error increases with increasing population size (Whitehouse et al. 2001, Morely & van Aarde 2007). Mark-recapture methods were therefore employed to study the demographics of the Majete elephant population through on-the-ground visual assessments and camera trapping. Results were used to determine the response of the population post reintroduction, providing a deeper understanding of reintroduction biology and thus contributing towards reserve management strategies.

Methods and Materials

Study Area Majete Wildlife Reserve is located at the southern tip of the Great Rift Valley in the lower Shire Valley region of southern Malawi (S15o 54’26.6”; E034o44’24.3”) (Figure 3.1.). It is a 700km2 reserve with northern and eastern boundaries defined by two perennial rivers, the Mkulumadzi and the Shire. The altitudinal gradient of the reserve varies, it is highest in the western region, Namisempha, where steeply undulating hills are distorted by river valleys and gradually decreases towards the eastern region of the reserve, the Sanctuary, where the terrain flattens towards the Shire River. Majete has two distinct seasons, the wet season (December to May) and the dry season (June to November). The expected annual precipitation in the reserve is between 680-800mm in the eastern lowlands and 700- 1000mm in the western highlands (Wienand 2013). Water availability is dependent on seasonal

Stellenbosch University https://scholar.sun.ac.za

rainfall, however, MWR has a number of perennial springs along with ten borehole-fed artificial waterholes. The vegetation is primarily woodland and it varies from low to high altitudes. The vegetation types are high altitude (410-770m) miombo woodland (Brachystegia boehmii, Burkea africana and Pterocarpus), medium altitude (230-410m) mixed woodland (Brachystegia boehmii, Diiospyrus kirkii and Combretum species), low altitude (205-280m) mixed woodland (Acacia species and Steculia) and savanna (Combretum species, Acacia species and Panicum species) (Patel, H. Unpub.).

Majete Wildlife Reserve was gazetted in 1955 but by 2003 most of its large game had been decimated due to poaching and poor management. In 2003, MWR underwent one of Africa’s greatest reintroduction programmes, after a Public Private Partnership (PPP) agreement was made between African Parks, Majete (Pty) Ltd. and the Malawian Department of National Parks and Wildlife (DNPW). Wildlife reintroductions to MWR were undertaken in stages as the reserve was not initially fenced in its entirety. The first stage involved fencing a smaller area of 140km2 in the north-eastern region, the Sanctuary, for reintroduction purposes in 2003. The second stage involved fencing the remaining boundary of the reserve and was completed in 2008, after which the sanctuary fence line was removed in 2011 and wildlife was able to roam into the greater reserve. In total over 2550 individual animals of 14 different species were reintroduced into MWR. Species reintroduced included elephant, black rhino, buffalo, sable, hartebeest and numerous other antelope species.

Stellenbosch University https://scholar.sun.ac.za

Methods Defining the population size, age and sex structure In order to determine the size, age structure and sex ratio of the elephant population in MWR the majority of individuals needed to be identified and so this study modified the methods used by Moss (1996), Whitehouse and Hall-Martin (2000), Moss (2001) and Morley and van Aarde (2007) to identify individuals and record their gender, age and other unique characteristics. Sampling was conducted using two methods. The first was field observations which comprised of randomised distance sampling and waterhole counts. The second was camera traps which recorded images of individuals and herds, where cameras were placed at various artificial waterholes and springs and along the road network in grids. In this particular section of the overall study, the population status was determined for 2015 and early 2016 and so field observations commenced in April 2015 and ended in April 2016 and camera trap images were recorded from January 2015 to January 2016. a) Individual Identification Techniques Using mark and re-sight methods, upon observation in the field photographs were immediately taken of elephants and individuals were ‘marked’ by recording various characteristics and unique markings which can be used to facilitate individual identification. These characteristics included for example: age, sex, body size, ear shape, any notches or holes present in the ears, patterns of blood vessels in the ears, wrinkles on the face, tusk size and configuration, lumps or scars on the body and kinks or baldness of the tail. Also recorded was the total number of individuals, GPS locations and numerous photographs of each individual were taken.

Elephants were later positively identified as ‘marked’ or ‘unmarked’ using a customised database, developed by the author (Appendix 2.1. & 2.2.). The database aided the identification of individual elephants as it contained photographs and profiles of all encountered individuals and herds, and information such as estimated age of individuals, sex and other identifying characteristics mentioned above were recorded. When an unknown animal was encountered it was given an identification code, it was sexed and aged and any unique characteristics were recorded, after which it was added to the ‘known’ population database (Appendix 2.3). After 13 months of sampling the majority of the elephants in the population were sexed and approximate ages estimated, based on size and developmental state. Since elephants grow throughout their lifetime, individuals were aged according to the following size categories (Moss 1996):

Infant (0-0.9 years): Infant’s shoulder was taller than the breast level of mother, reaching wrinkles above elbow.

Stellenbosch University https://scholar.sun.ac.za

Calf (1-4.9 years): Top of calf’s shoulder above mother’s armpit, back was level with anal flap and reached lower quarter of mother’s ear. Tusks 5-7cm.

Juvenile (5-9.9 years): Overall size of juvenile was about three quarters of adult female. Tusks splayed and 25-30cm in length.

Small Adult Female (10-19.9 years): Same size as other adult females. Tusks had adult configuration (convergent, straight or asymmetrical). Body was square in shape compared to older females who were more rectangular.

Small Adult Male (10-19.9 years): Same size or slightly taller than adult females. Head shape (sloping rather than angular) was more noticeable, tusk circumference was thicker than females of same age.

Medium Adult Female (20-34.9 years): Circumference of tusks at base was distinctly bigger than small adult females. Bodies became more rectangular.

Medium Adult Male (20-34.9 years): Taller than all adult females, head began to thicken and change into an hour glass shape, wide at the eyes and the base of the tusks.

Large Adult Females (>35 years): Tusks marginally thicker, back lengthened so animal appeared long in body. Older females hollowed above the eyes, ears held lower, longer back length.

Large Adult Males (>35 years): Very big, towered over largest females. Neck thick, overall body heavy set, tusk circumference at lip greater than younger males. b) Distance Sampling Established roads were used to conduct drive transects throughout MWR. Sampling in the Sanctuary was conducted on a weekly basis and sampling in the remainder of the reserve was conducted on a monthly basis due to logistics. Sampling generally began just before dawn and continued until after sunset, with a rest period during the hottest part of the day. The majority of sampling was conducted by a single observer, however, from June to December 2015 volunteers assisted in the observation and recording of data during counts. An average cruising speed of 15 kilometres per hour was maintained when searching for animals and due to the dense nature of the vegetation in the reserve, visibility of elephant was generally less than 150m. Elephants were initially sighted with the naked eye and binoculars were used when necessary to determine age, sex and observe other unique characteristics. Data recorded for distance sampling included: date, time of observation, GPS position of observer, total number of elephant seen, number of males and females and the respective age classes of individuals (infant, calf, juvenile, small adult, medium adult and large adult). Additionally, photographs were taken at each observation using a Nikon D40x DSLR camera. Images were later

Stellenbosch University https://scholar.sun.ac.za

analysed and added to the database. Images recorded from distance sampling were analysed using the same method as described above. c) Waterhole Counts MWR has ten borehole-fed artificial water points and the majority of these water points were powered by solar energy. Counts were conducted at four artificial water points in the Sanctuary (Nsepete, Nakamba, Thawale and the Heritage) and one artificial water point in the Pende region (Nthumba) from June to December 2015. Counts were conducted for 12 hour periods and commenced at 06h00 and ended at 18h00. In the Sanctuary, three to four observers were stationed on a viewing platform or in an elevated hide from which elephants were monitored at the waterhole. Data recorded included the weather conditions (cloud cover, temperature), total number of elephants sighted, time of observation of individual or herd, the gender and ages of each individual and several behavioural observations (interspecific or intraspecific interactions). Additionally, photographs of each sighting were taken with a Nikon D40x DSLR camera and later added to the database. d) Camera Trapping Camera trapping, a non-invasive research technique, has been used in a number of studies to determine the relative abundance of certain species across a specific area. In MWR camera traps were a useful form of sampling as they increased the sampling area and allowed for continued monitoring of sites that, due to time constraints, could not be accessed more than once a month. In this study, CuddebackTM Ambush© Black Flash© camera traps were stationed at specific artificial waterholes throughout the reserve for one dry season (June 2015 - November 2015) and one wet season (December 2015 - April 2016). Cameras were placed at seven of the artificial waterholes and at one perennial spring (Pwadzi Spring) due to the ease of access of these water points. Due to maintenance and technical issues at two waterholes (Pende 1 and Nthumba) only data from five water points (Nsepete, Nakamba, Diwa, Pende 2 and Pwadzi spring) was used in this study. Each camera was placed at each station for a period of 30 days after which SD cards and batteries were changed. Once images were downloaded from SD cards they were sorted, according to date and location, and saved into the appropriate files on an external hard drive.

In order to identify elephant individuals and herds in MWR individual identification techniques (as mentioned above). All data recorded from images was entered into Microsoft Excel 2013 and the customised database. In terms of analysing the camera trap images the following assumptions applied: firstly a single image containing elephants (bulls or family herds) was considered a single observation if no other photographs were taken. Secondly, if individuals were captured on camera in a continuous sequence of photographs in a short period of time (approximately 20 min) on the same day they were assumed to be from the same herd. Thirdly, when family herds were confirmed the

Stellenbosch University https://scholar.sun.ac.za

sequence of photographs captured of the herd within 60 min was considered a single observation for that specific herd. Fourthly, once an extended period of time, longer than 60 min, had passed between photographs and new individuals had been confirmed they were assumed to be from a different herd. Lastly, small, medium and large adult bulls were assumed to have left their original family herds. This is as bulls tend to leave their family herds between the ages of 11 and 13 years. If an individual bull was captured on a photograph with a family herd, notes were taken but the individual was assumed to range alone. e) Aerial census In order for MWR management to determine the total size of the elephant population in the reserve aerial counts were conducted in 2010, 2012 and 2015. All aerial count data was provided by African Parks (Pty) Ltd. Exact methods and aerial count teams differed over the years, however, the basic procedure was the same. All aerial counts were conducted in the dry season and each aerial count consisted of a pilot and 1-3 observers. Counts were flown in transects, the calibrated strip width of each transect was 500m and flight paths were orientated in an east to west direction. Due to the size of MWR counting blocks were established (blocks varied in size and location per aerial count), each block in the reserve was counted on a separate day and all blocks were counted on consecutive days. When animals were sighted data recorded included the GPS track log of all transects, game species and number of individuals, and any significant waypoints were also marked using a GPS.

Comparing the current population with populations from the initial reintroductions In order to determine how much the current elephant population has changed demographically since the reintroductions into MWR in 2006, 2008 and 2010, the data gathered from this study were compared to the capture records from the three translocation phases. All data regarding the capture records were acquired from the DNPW. However, the capture records from the 2006 and 2008 translocations were missing, so as a result only the 2010 capture data were utilised in this study. The capture records contained information such as the total number of individuals and the gender and ages of each individual translocated to Majete from Liwonde National Park and Pirilongwe Forest Reserve.

Habitat use and movement The primary method used to determine the elephants’ habitat use and movement in MWR was camera trapping. Sampling employed the same camera trap method mentioned above, where by camera traps were placed at certain perennial water points in MWR. Unlike in the previous section, this study required data from a number of different years in order to observe how habitat use and movement possibly changed over time. Thus, camera traps were placed at perennial water sources in

Stellenbosch University https://scholar.sun.ac.za

2014, 2015 and 2016. To sample elephant’s waterhole use in different habitats and at different altitudes camera traps were stationed at Nakamba, Nsepete, Pende 2, Diwa and Pwazi spring (Figure 3.1). SD cards and batteries were changed on a monthly basis and all images were analysed using Timelapse Image Analyser 2.0 (Greenberg Consulting Inc. 2016).

Camera trap gridlines were established in MWR to determine elephant movement over time and to observe whether elephants influence the presence of other species. The camera trap grid delineated the 700km2 reserve into four grid lines running from north to south. Sampling took place in the wet (January - May) and the dry (June - October) seasons in 2014 and 2015. All grid lines were designed according to the road network in MWR and covered the four different vegetation types, namely, high altitude woodland, medium altitude woodland, low altitude woodland and savanna. Only one of the grid lines was placed out in the field at any given time for a period of 30 days and once one grid was retrieved the next grid was immediately positioned. Each grid line consisted of eight stations and each station had a single CuddebackTM Ambush© Black Flash© camera attached to a specific tree/shrub. Cameras were placed at each gridline once per season and twice per year (wet and dry seasons) and the same protocol was followed for 2014 and 2015. Stations were preferentially placed at jeep track intersections, wildlife trails, or in river beds. Before all cameras were placed in the field the time and date settings were checked and cameras were set to take photos at one minute intervals once motion sensors were triggered. A total of 32 camera stations were set out for the period of this study and data was analysed from 30 of these stations

For this particular section, all images acquired from camera traps were analysed in Timelapse Image Analyser 2.0 (Greenberg Consulting Inc. 2016) where the time, date, species, total number of individuals, gender and ages were recorded. In order to determine whether elephants influenced the presence of other species in the reserve, images recorded from camera grids were analysed in terms of all species that occur in MWR and not just elephants. Images recorded from camera traps set at water points, only included elephants so as to determine which water points were preferred and how family herds and bulls differed in their water point usage. The assumptions mentioned above apply to this section but for clarification, an observation at a water point was defined as the presence of elephants (bulls or herds) at a water point for a period of no shorter than 5 minutes and no longer than an hour. Once the time frame was over, unless proven otherwise, a new observation was started with a new elephant. All data from Timelapse Image Analyser 2.0 (Greenberg Consulting Inc. 2016) were exported into Microsoft Excel 2013, sorted and later analysed. Lastly, all mapping displaying data from camera grids as well as water points was conducted in QGIS 2.16.0 (Creative Commons Attribution-ShareAllike 3.0 Licence (CC BY-SA)).

Stellenbosch University https://scholar.sun.ac.za

Statistical Analysis All statistical analyses were performed using Statistica, version 13 (©Statsoft Inc., 2016). To determine whether there was any significant statistical change in firstly, the size of the elephant population and secondly, the age structure of the population from 2006 to 2015, a regression analysis was used. To determine whether there was any significant difference between the presence of elephant in MWR in different seasons (wet or dry) and years (2014 and 2015) a Mann-Whitney U Test was used. In addition, a factor analysis, the varimax normalised rotation, and a cluster analysis was used to determine whether there was any correlation between the presence of elephants and other herbivore species. Lastly, in order to determine if waterhole usage by the elephant population (family herds and bulls) differed significantly on a seasonal basis a chi-squared test and an ANOVA with a bootstrap multiple comparisons analysis were used. Lastly, all mapping was completed using QGIS, version 2.16 (Creative Commons Attribution-ShareAlike 3.0 Licence) and shapefiles were provided by African Parks (Pty) Ltd.

Results

To determine current population size, age and sex structure

The 2015 aerial census counted a total of 389 individual elephants throughout MWR. Using mark- recapture methods the current study identified a total of 366 individuals of these individuals from January 2015 to April 2016 (Table 3.1.). Within the 366 individuals identified, 47 different family herds were identified as well as 91 solitary bulls (Appendix 2.3). The age structure of the population composed of 43 infants, 87 calves, 53 juveniles, 55 small adults, 98 medium sized adults and 30 large adults (Table 3.1.). The largest age class portion of the population was the medium adults (26.8%), followed by calves (23.8%) (Table 3.1., Figure 3.2.). Sex ratios for individuals under the age of 10 years was even at a ratio of 1:1, however, there were 84 individuals of unknown gender. The sex ratio for individuals between the ages of 10-60 years was relatively even at 1:1 (Table 3.1.; Figure 3.2.).

Stellenbosch University https://scholar.sun.ac.za

Age Class Unknown Age Class Total Males Females Sex Ratio Proportion of Sex Population (%)

Infant (<1 year) 43 5 5 33 11,75 52:47 Calf (1-4.9 years) 87 26 19 42 23,77 (1.1:1) Juvenile (5-9.9 years) 53 21 23 9 14,48

Small Adult (10-19.9 years) 55 33 22 15,03 Medium (20-34.9 years) 98 48 50 91:92 (1:1) 26,78 Adult Large Adult (>35 years) 30 10 20 8,20

Total 366 143 139 84 100.00%

8% 12%

Infant (<1 yr) Calf (1-4.9 yr) 27% 24% Juvenile (5-9.9 yr) Small Adult (10-19.9 yr) Medium Adult (20-34.9 yr) Large Adult (>35 yr)

15% 14%

Figure 3.2. The proportional representation of each age class in the elephant population on MWR, Malawi, in the years 2015/2016.

Stellenbosch University https://scholar.sun.ac.za

To determine the demographic performance of the population since reintroduction

Since the reintroduction of elephant into MWR in 2006, 2008 and 2010 the population has increased significantly to an estimated total of 389 individuals (Figure 3.3.)(p=0.00; r2=1). With minimal mortality of individuals, the six years from the conclusion of reintroductions in 2010 to the aerial count in 2015, saw an annual population growth rate of 13.8%. In terms of the change in population structure over time, there was a significant increase in the number of adult males (p=0.04; r2=0.8) and adult females (p=0.01; r2=0.9) from 2006 to 2015. However, due to the lack of data from the 2006 and 2008 reintroductions, statistical analysis on the younger age groups, namely juveniles and calves, as well as a more detailed analyses was not possible. This is as the current data set does not contain any data regarding gender and age classes for elephants’ pre the 2010 reintroduction.

Figure 3.3. The cumulative total number of elephant in MWR since the start of the reintroduction process in 2006. (p=0.00007, r2=0.9971)

To determine habitat use and movement

The movement patterns of the presence/absence of elephant on the reserve differed significantly in the wet and dry seasons (p=0.01) but there was no significant difference from year to year (p=0.5). In the wet seasons elephants appeared to expand their movement to the western side of the reserve that has few perennial water sources and is at a higher altitude (Figure 3.4.). In the dry seasons elephants appeared to remain close to perennial water sources such as the rivers, waterholes and

Stellenbosch University https://scholar.sun.ac.za

springs on the eastern side of the reserve (Figure 3.4.). In both the dry and wet seasons it is clear that elephants prefer the north eastern area of the reserve that is close to the Shire and Mkulumadzi Rivers but a high number of individuals were also recorded in the south eastern parts of the reserve (Figure 3.4.). The map produced from the 2015 aerial survey, which was conducted in the dry season, demonstrated greater movement in the western part of the reserve and less movement in the south west, compared to the camera grids (Figure 3.5.). However, there was still a large number or elephants in the north eastern section of the reserve where the Shire and Mkulumadzi Rivers are located (Figure 3.5.).

Figure 3.4. Maps representing the elephant presence across MWR in two different wet and dry seasons in relation to altitude and presence of perennial water sources. Number of individuals recorded at each site is correlated to the size of the circular marker.

Stellenbosch University https://scholar.sun.ac.za

Figure 3.5. Map representing the elephant group number and position in MWR in the 2015 aerial survey. (per comms. African Parks (Pty) Ltd.)

To test whether the presence of elephant impacted other species around MWR, a factor analysis was used. This was done as a preliminary test to investigate whether the presence of elephant influenced the presence of other high density herbivore species on MWR such as waterbuck and buffalo. Elephant presence was found to be closely correlated to the presence of warthog, impala and waterbuck in both the wet and dry seasons (Figure 3.6.). Elephant presence was found to be weakly correlated to vervet monkeys, nyala and baboon in both the wet and dry seasons (Figure 3.6.).

Figure 3.6. A joint tree cluster analysis representing the correlation between the presences of various species in MWR.

Stellenbosch University https://scholar.sun.ac.za

In general, there was a significant difference between elephant water point usage in the wet and dry seasons, as a greater number of observations were recorded in the dry when compared to the wet seasons (p=0.00; χ2=216.61; df=5). More specifically the water points showing the highest number of observations in the wet season of 2014 were Diwa and Pende 2. In the dry season of 2014, Pende 2 and Pwadzi Spring had the highest number of observations. In the dry season of 2015 it was Pwadzi spring and Nsepete and lastly in the wet season of 2016 it was Pwadzi spring (Figure 3.7.). There was also a significant difference of water point usage between bulls and family herds in the different years and across seasons (p=0.00). In general, there were a higher number of observations of family herds at water points in both the dry and wet seasons compared to bulls. In the wet season of 2014, family herds dominated the usage of all the water points, although there appeared to be more bulls at Diwa and Pende 2 water points compared to the rest (Figure 3.7.). In the dry season of 2014, no data was available for Diwa (due to maintenance) but it appeared that family herds dominated the use of the water points and the highest number of observations were recorded at Pende 2 (Figure 3.7.). In the dry season of 2015 family herds again dominated water point usage and Pwadzi spring recorded the highest number of observations (Figure 3.7.). Lastly, the wet season of 2016 bulls appeared to dominate Diwa, Nsepete and Nakamba water points while the family herds appeared to dominate Pende 2 and Pwadzi spring and the highest number of observations were recorded at Pwadzi spring (Figure 3.7.).

Stellenbosch University https://scholar.sun.ac.za

Figure 3.7. Maps representing elephants, more specifically bulls and family herds, water point usage in MWR in the wet and dry seasons of 2014, 2015 and 2016. The size of the pie chart is correlated to the number of observations at each water point.

Discussion

Population Status and Growth Rate

The most recent aerial count conducted at MWR counted a total of 389 individual elephants in 2015 and the current study was able to formally identify 366 of these individuals, using mark-recapture techniques, by April 2016. It has been established that aerial counts tend to under sample areas, more specifically woodland areas, and so it is proposed that the actual population size of elephant is larger than 389 and is more likely closer to 430 individuals (Caro 1999, Walsh et al. 2001, Whitehouse et al. 2001, Payne et al. 2003). In the absence of migration and heavy poaching, the Majete elephant population has grown at an estimated annual rate of 13.8% since the completion of reintroductions in 2010. The growth rate of the population will not have been uniform over time but given the current level of protection and the available range, it is predicted that the population will continue to increase.

Other African elephant populations, specifically those that are unfenced and therefore exposed to human-elephant conflict and poaching, such as the ANP and SNR populations, have been reported to have average population growth rates of only 2.2% and 0.17%, respectively (Moss 2001; Wittemyer et al. 2014). These low growth rates were, amongst other factors, attributed to the aging population in ANP and the migration in SNR (Moss 2001; Wittemyer et al. 2014). However, in fenced reserves

Stellenbosch University https://scholar.sun.ac.za

with resource provision and minimal human-elephant conflict, including poaching, such as in AENP the population growth rate averaged 6.0% (Whitehouse & Hall Martin 2000). This suggests that increased protection and resource provision were the main contributors to the difference in population growth rates between ANP, SNR and AENP. However, MWR’s population growth rate is far greater than these other populations indicating that resource provision and protection may not be the only factors contributing to the population growth rate in Majete.

Calef (1988) predicted that the maximum rate of increase for elephant populations is 7.0% per year. This maximum rate of increase is associated with optimal conditions, such as a small population that has been introduced into a new habitat that previously had no elephant, or a population that has been reduced to levels well below its carrying capacity (Calef 1988). A possible reason for the extremely high population growth rate on Majete is that the elephant population is experiencing optimal conditions required for maximum growth rate. When a population experiences these conditions, mortality is extremely low and fecundity approaches the physiological maximum, which suggests that the calving interval would be 22 months (one gestation length) after a birth if the first calf dies or 27 months after the birth of a surviving calf (Calef 1988; Moss 2001). This, however, cannot be sustained and the population will begin to stabilise with a stable age distribution of 48% of individuals older than 11 years, which is the average age of first reproduction, and 6.7% of infants (<1year) (Calef 1988). Elephant populations such as those in AENP have therefore already stabilised to a certain extent, hence their lower population growth rate. As the MWR population has an extremely high growth rate it can be argued that the population is operating under optimal conditions and has not yet stabilised.

This theory is further supported by the fact that the Majete population structure does not match the description of a population that has stabilised. The Majete population structure consists of 11.8% infants (<1 year) and 50% of individuals above the age of 11 years, which as mentioned is the average age of first reproduction. With such a high proportion of infants, a sex ratio of 1:1 in both adults and young and a minimal mortality rate, it is assumed that the population will continue to increase before eventually stabilising. According to Bothma & du Toit (2010) a rough estimate of the recommended ecological capacity of MWR, based on the average annual rainfall and the size of the reserve, is an estimated 304 individuals. Thus, the ecological capacity on MWR, specifically with regards to elephants, has already been surpassed and the population should begin to stabilise in the not too distant future due to increasing pressure on the vegetation. However, it must be acknowledged that instantaneous counts such as this one, (mark-recapture over a period of a year), could merely reflect the short term patterns of ‘highs’ and ‘lows’ and could potentially produce misleading estimates of the population’s actual growth rate. In order to avoid this, population counts should be conducted every few years and not just rely on a single count.

Stellenbosch University https://scholar.sun.ac.za

Habitat Use and Movement

In general, the African elephant’s dependence on surface water means that the availability and distribution of water sources strongly influences the species dispersal in a landscape (Chamaille- Jammes et al. 2008; Smit & Ferreira 2010). The camera grids that recorded species presence and absence around MWR suggested that the elephants preferred areas around water sources in both the dry and wet seasons. More specifically the highest number of individuals recorded in both seasons was in the Sanctuary, where the Shire and Mkulumadzi Rivers, one spring and four artificial waterholes were located. In the wet seasons a larger number of individuals were recorded in Pende, the south east region, compared to the dry season. However, the number of individuals recorded in Pwadzi, the south western region, and Namisempha, the north western region, increased slightly in the dry season. This is indicative of seasonal movement and in regions with a distinctly seasonal rainfall, such as MWR, regional elephant distribution during the dry season is dictated by surface water availability and location (Stokke & du Toit 2002; Shannon et al. 2005; Chamaille-Jammes et al. 2007; de Beer & van Aarde 2008, de Kneght et al. 2011). In the wet season elephants disperse widely across the landscape due to the higher availability of surface water through ephemeral pools and rivers (Stokke & du Toit 2002; Shannon et al. 2005; Chamaille-Jammes et al. 2007; de Beer & van Aarde 2008, de Kneght et al. 2011).

When comparing the camera grid maps to the aerial count map it is clear there was a lot of elephant movement that the camera grids did not record. This was probably due to the camera traps being restricted to the road network within the reserve, whereas the aerial count was able to record data throughout the reserve. The aerial count was conducted in the peak of the dry season in 2015, so it was anticipated that elephant movement would be restricted to perennial water sources mainly in the Sanctuary and Pende regions. Yet, the aerial count suggests that there was a lot more movement of elephants in the Diwa region than initially anticipated, an area that is water scarce. This suggests two things, firstly there may be additional water sources, in the form of springs, in the central and western regions. Secondly, both bulls and family herd’s home ranges may be quite large and may span the majority of the reserve rather than just a few regions. This is supported by the fact that one of the family herds that was thought to reside in Pende was recorded at Diwa, the artificial waterhole that is located at the highest altitude in the Namisempha region, in the wet season (author, unpublished data). Additionally, another family herd was recorded in both the Sanctuary as well as Pende in the dry season (author, unpublished data). This indicates that there may be more movement occurring than initially anticipated. Additional studies using GPS collars would need to be conducted in order to determine the exact ranging behaviour of the elephant in MWR.

Stellenbosch University https://scholar.sun.ac.za

Elephants respond to their environment in a scale-dependent manner, at a coarse spatial scale forage characteristics drive habitat selection and at finer spatial scales surface water drives habitat selection (de Knegt et al. 2011). Habitat use generally increases with proximity to water sources and factors such as season and the nature of the surrounding habitat (ie: vegetation type, terrain or altitude) affect water point usage (Harris et al. 2008; Gaugris et al. 2009; Staub et al. 2013; Weinand 2013). The analyses indicated that waterhole usage in MWR changed seasonally, similar to the results obtained by Weinand (2013). This was confirmed as there was less usage of the selected water sources in the wet season when compared to the dry season. The waterholes that were used the most were Pende 2 and Pwadzi Spring. Weinand (2013) and other previous studies show that this is because both these water points are situated at lower altitudes, usually favoured by elephants, and both water points are surrounded by the preferred vegetation types of low altitude woodland and riparian vegetation (de Knegt et al. 2011; Staub et al. 2013). Staub et al. (2013) documented higher browsing levels in vegetation communities at lower altitudes in MWR and elephants primarily favoured riparian woodland, followed by Acacia-dominated woodland and lastly Brachystegia-dominated woodland. Naturally, the higher water point usage at Pende 2 and Pwadzi indicates that elephants prefer these two areas.

The water points in the Sanctuary, Nsepete and Nakamba, were amongst the least used perennial water points in MWR. This was unexpected as Weinand (2013) found that these water points were amongst the most used by elephants. A possible reason for the decrease in water point usage at the time of this study, is that the Sanctuary has the highest density of perennial water sources in the reserve and so elephant water point usage is spread out across the region and is not focused on these two artificial waterholes. This is supported by the likely fact that the piospheres in the Sanctuary have merged due to the closeness of perennial water points which are all within 5km of each other (Weinand 2013). Therefore, as elephants prefer to use habitat within 4km of water throughout the year (de Beer et al. 2006), distance does not limit elephant water point usage in the Sanctuary. In addition, the merging of the piospheres around water points in the Sanctuary suggests that elephant habitat use and water point use are both homogenous in the region.

In addition, in the study by Weinand (2013) there was a relatively high usage of Diwa waterhole compared to Nakamba and Nsepete waterholes. This result was unexpected as Diwa is at a higher altitude, is surrounded by undulating terrain and Brachystegia-dominated woodland, all of which elephants least prefer (Knegt et al. 2011; Staub et al. 2013). A possible reason for this result is that the elephant population in MWR has significantly increased since 2013 and so due to a higher density of elephant, home ranges and habitat use are now expanding into least preferred altitudes and vegetation types (Junker et al. 2008). This is supported by Chamaille-Jammes et al. (2007) who found

Stellenbosch University https://scholar.sun.ac.za

that there was less of a difference in water point usage with elephants at a higher population density, compared to elephants at lower population densities. This is due to water points reaching their full capacity or becoming saturated with elephants at a higher density. This supports the aerial count and mark-recapture results, which indicate that the elephant population has increased dramatically to the point that they are now beginning to utilise less preferred regions of the reserve due to higher densities.

Lastly, there was an indication from the analyses that there was a significant difference between family herd and bull waterhole use in the wet and dry season. Family herds tended to dominate waterhole use in both seasons, apart from in the wet season in 2016 where bulls appeared to dominate waterhole usage in both the Sanctuary as well as at Diwa waterhole. Bulls are able to adapt their movement behaviour according to forage quality and abundance (Stokke & du Toit 2000; Woolley et al. 2009). Smaller bodied females, however, cannot exhibit the same flexibility as they have higher nutritional demands and the energetic constraints of group living with juveniles and calves (Stokke & du Toit 2000; Woolley et al. 2009). Additionally, in the dry season family units tend to remain close to perennial water sources, staying within 3.5km from water sources, compared to bulls who roam as far as 10km from perennial drinking water in the dry season (Stokke & du Toit 2002). In the wet season, due to ephemeral pools, all groups tend to average 5km from water points (Stokke & du Toit 2002). It is, therefore, possible in MWR that waterholes are dominated by family groups in the dry season due to their localised home ranges but bulls are able to roam widely and so spread their waterhole use throughout the reserve. The fact that bulls dominated the Sanctuary in the wet season of 2016 suggests that either family units were utilising other water sources in the Sanctuary or they expanded their home ranges into other areas of the reserve due to the abundance of water in the wet season.

Conclusion

In the absence of migration and heavy poaching the Majete elephant population has grown at an estimated annual rate of 13.8% since the completion of the reintroduction programme in 2010. The estimated size of the population for 2015/2016 is more than likely larger than 389 individuals. The population has an extremely high growth rate and it is postulated that it is operating under optimal conditions and has not yet begun to stabilise. This theory is supported by the population structure as it consists of 11.8% infants (<1 year) where a stable population would only consist of 6.7% infants (Calef 1988). With such a high proportion of infants, a sex ratio of 1:1 in both adults and young and a minimal mortality rate it is assumed the population will continue to increase but should stabilise in the near future.

Stellenbosch University https://scholar.sun.ac.za

In order to manage the current elephant population in MWR and to prevent any adverse effects on the environment, such as woody species loss, it is suggested that the population either stabilises or reduces its growth rate. In accordance with the current management strategy in MWR, the recommended management options are translocation and contraception (van Aarde & Jackson 2007). The current elephant population is already above the recommended ecological capacity, reducing the number to below the ecological capacity through translocation and thereafter implementing contraception on remaining herds will ultimately slow the growth rate (van Aarde & Jackson 2007). Long-term monitoring programmes of the elephants should be implemented in order to monitor the population’s demographics and size. Recording cumulative data over time will provide important demographic information, for example, the average age females first give birth and the average calving interval of females, which is extremely useful in elephant management (Whitehouse and Hall- Martin 2000; Moss 2001). Furthermore, as recommended in Weinand (2013) long term monitoring programmes of elephants’ impact on woody vegetation due to surface water availability need to be implemented especially in a fenced reserve such as MWR. The combination of population control measures (translocation and contraception), water point management and monitoring will provide the majority of information needed to develop a sufficient elephant management plan for MWR.

Acknowledgements

Thank you to Dr Alison Leslie for your advice and guidance. Thank you to Professor Dan Nel for the statistical help. Thanks to Aileen Thompson and Claire Cordon for their advice regarding QGIS. To the DNPW, thank you for your cooperation in searching for the capture records. To all those wonderful souls (Alison L, Allison C, Maya, Claire, Leslie, Fafa, Charli, Willem, Heather, Nina and Barbra) who helped me collect data and take photos of countless elephants, thank you so very much you are the best! A big thanks to the Earthwatch Institute for funding the Animals in Malawi Project and for funding this research. Lastly, thank you to the Earthwatch volunteers for helping me in the field, your passion and enthusiasm was inspiring and your photographs were most helpful.

Stellenbosch University https://scholar.sun.ac.za

References

 Barnes, R.F.W. (1996) Estimating forest elephant abundance by dung counts. In: Studying Elephants. AFW Technical Handbook, Series 7 (Ed.K.Kangwana). African Wildlife Foundation, Nairobi.  Blake, S., Deem, S.L., Mossimbo, E., Maisels, F. & Walsh, P. (2009) Forest Elephants: Tree Planters of the Congo. Biotropica, 41, 459–468.  Bothma, J. du P. & du Toit, J.G. (2010) Game Ranch Management, 5th edn. Van Schaik Publishers, Pretoria, South Africa.  Bouche, P., Douglas-Hamilton, I., Wittemyer, G., Nianogo, A.J. & Doucet, J.L. (2011) Will Elephants Soon Disappear from West African Savannahs? Plos One, 6, e20619.  Buss, I. O. & Smith, N. S. (1966) Observations on reproduction and breeding behaviour of the African elephant. Journal of Wildlife Management, 30, 375-388.  Calef, G.W. (1988) Maximum rate of increase in the African Elephant. African Journal of Ecology, 26, 323-327.  Caro, T.M. (1999) Conservation monitoring: estimating mammal densities in wooded habitats. Animal Conservation, 2, 305–315.  Chamaille-Jammes, S., Fritz, H., Valeix, M., Murindagomo, F. & Clobert, J. (2008) Resource variability, aggregation and direct density dependence in an open context: the local regulation of an African elephant population. Journal of Animal Ecology, 77, 135-144.  Chamaille-Jammes, S., Valeix, M. & Fritz, H. (2007) Managing heterogeneity in elephant distribution: interactions between elephant population density and surface water availability. Journal of Applied Ecology, 44, 625-633.  Corfield, T. F. (1973) Elephant mortality in Tsavo National Park, Kenya. African Journal of Ecology, 11, 339-368.  Croze, H. (1972) A modified photogrammetric technique for assessing age structures of elephant populations and its use in Kidepo National Park. African Journal of Ecology, 10, 91- 115.  de Beer, Y. & van Aarde, R.J. (2008) Do landscape heterogeneity and water distribution explain aspects of elephant home range in Southern Africa’s arid savannas?. Journal of Arid Environments, 72, 2017-2025.  de Knegt, H.J., van Langevelde, F., Skidmore, A.K., Delsink, A., Slotow, R., Henley, S., Bucini, G., de Boer, W.F., Coughenour, M.B., Grant, C.C., Heitkonig, I.M.A., Henley, M., Knox, N.M., Kohi, E.M., Mwakiwa, E., Page, B.R., Peel, M., Pretorius, Y., van Wieren, S.E. & Prins, H.H.T. (2011)

Stellenbosch University https://scholar.sun.ac.za

The spatial scaling of habitat selection by African elephants. Journal of Animal Ecology, 80, 270-281.  Douglas-Hamilton, L. (1996) Counting elephants from the air-total counts. In: Studying Elephants. AFW Technical Handbook, Series 7 (Ed.K.Kangwana). African Wildlife Foundation, Nairobi.  Dublin, H.T., Sinclair, A.R.E. & McGlade, J. (1990) Elephants and fire as causes of multiple stable states in the Serengeti Mara woodlands. Journal of Animal Ecology, 59, 1147–1164.  Gaillard, J.M., Festa-Bianchet, M. & Yoccoz, N.G. (1998) Populations dynamics of large herbivores: variable recruitment with constant adult survival. Trends in Ecology & Evolution, 13, 58–63.  Gaillard, J.M., Festa-Bianchet, M., Yoccoz, N.G., Loison, A. & Toigo, C. (2000) Temporal variation in fitness components and population dynamics of large herbivores. Annual Review of Ecology and Systematics, 31, 367–393.  Gaugris, J.Y. & van Rooyen, M.W. (2009) Effects of water dependence on the utilisation pattern of woody vegetation by elephants in the Tembe Elephant Park, Maputaland, South Africa. African Journal of Ecology, 48, 126-134.  Gough, K.F. & Kerley, G.I.H. (2006) Demography and population dynamics in the elephants Loxodonta africana of Addo Elephant National Park, South Africa: is there evidence of density dependent regulation? Oryx, 40(4), 434-441.  Hall-Martin, A. (1992) Distribution and status of the African elephant, Loxodonta africana, in South Africa, 1652-1992. Koedoe, 35(1), 65±88.  Hanks, J. (1969) Seasonal breeding of the African elephant in Zambia. African Journal of Ecology, 7, 167.  Hanks, J. (1972) Reproduction of the elephant (Loxodonta africana) in Luangwa Valley, Zambia. Reproduction, 30, 13-26.  Harris, G.M., Russeli, G.J., van Aarde, R.I. & Pimm, S.L. (2008) Rules of habitat use by elephants Loxodonta africana in Southern Africa: insights for regional management. Oryx, 42 (1), 66-75.  Jachman, H. (1991) Evaluation of four survey methods for estimating elephant densities. African Journal of Ecology, 29, 188-195.  Jachman, H. (1996) Direct counts from the ground. In: Studying Elephants. AFW Technical Handbook, Series 7 (Ed.K.Kangwana). African Wildlife Foundation, Nairobi.  Jachmann, H. & Bell, R. H. V. (1984) The use of elephant droppings in assessing numbers, occupancy and age structure: a refinement of the method. African Journal of ecology, 22, 127- 141.

Stellenbosch University https://scholar.sun.ac.za

 Junker, J., van Aarde, R.J. & Ferreira, A.M. (2008) Temporal trends in elephant Loxodonta africana numbers and densities in northern Botswana: is the population really increasing? Oryx, 42(1), 58-65.  Kerley, G.I.H. & Landman, M. (2006) The impacts of elephants on biodiversity in the Eastern Cape Subtropical Thickets. South African Journal of Science, 102, 395-402.  Laws, R. M. (1966) Age criteria for the African elephant, Loxodonta africana. African Journal of Ecology, 4, 193-217.  Laws, R.M. (1970) Elephants as Agents of Habitat and Landscape Change in East-Africa. Oikos, 21, 1-15.  Leuthold, W. (1976) Age structure of elephants in Tsavo National Park. Journal of Applied Ecology, 13, 435-444.  Lewis, D.M. (1984) Demographic changes in the Luangwa Valley Elephants. Biological Conservation, 29, 7-14.  Lindeque, M. (1991) Age structure of the elephant population in the Etosha National Park, Namibia. Madoqua, 18, 27-32.  Loarie, S.R., van Aarde, R.J. & Pimm, S.L. (2009) Fences and artificial water affect African savannah elephant movement patterns. Biological Conservation, 142(12), 3086-3098.  Mbugua, S. (1996) Counting elephants from the air - sample counts. In: Studying Elephants. AFW Technical Handbook, Series 7 (Ed.K.Kangwana). African Wildlife Foundation, Nairobi.  Morley, R.C. & van Aarde, R.J. (2007) Estimating abundance for savanna elephant population using mark-resight methods: a case study for the Tembe Elephant Park, South Africa. Journal of Zoology, 271, 418-427.  Morrison, T.A., Holdo, R.M. & Anderson, M. (2016) Elephant damage, not fire or rainfall, explains mortality of overstorey trees in Serengeti. Journal of Ecology, 104, 409-418.  Moss, C. J. (1996) Getting to know a population. Studying elephants, African Wildlife Foundation Technical Handbook Series 7: 58-74. Kangwana, K. (Ed.). Nairobi: AWF.  Moss, C.J. (2001) The demography of an African elephant (Loxodonta africana) population in Amboseli, Kenya. Journal of Zoology, 255, 145-156.  Owen-Smith, N., Kerley, G.I.H., Page, B., Slotow, R. & van Aarde, R.J. (2006) A scientific perspective on the management of elephants in the Kruger National Park and elsewhere. South African Journal of Science, 102, 389–394.  Owen-Smith, N., Mason, D.R. & Ogutu, J.O. (2005) Correlates of survival rates for 10 African ungulate populations: density, rainfall and predation. Journal of Animal Ecology, 74, 774–788.

Stellenbosch University https://scholar.sun.ac.za

 Payne, K.B., Thompson, M. & Kramer, L. (2003) Elephant calling patterns as indicators of group size and composition: the basis for an acoustic monitoring system. African Journal of Ecology, 41, 99–107.  Power, M.E., Tilman, D., Estes, J.A., Menge, B.A. & Bond, W.J. (1996) Challenges in the quest for keystones. Bioscience, 46, 609–620.  Pringle, R.M. (2008) Elephants as agents of habitat creation for small vertebrates at the patch scale. Ecology, 89, 26–33.  Shannon, G., Page, B., Slotow, R. & Duffy, K. (2006) African elephant home range and habitat selection in Pongola Game Reserve, South Africa. African Zoology, 41 (1), 37-44.  Short, R. V. (1966) Oestrous behaviour, ovulation and the formation of the corpus luteum in the African elephant, Loxodonta africana. African Journal of Ecology, 4, 56-68.  Skarpe, C., Aarrestad, P.A., Andreassen, H.P., Dhillion, S.S., Dimakatso, T., du Toit, J.T., Halley, D.J., Hytteborn, H., Makhabu, S., Mari, M., Marokane, W., Masunga, G., Modise, D., Moe, S.R., Mojaphoko, R., Mosugelo, D., Motsumi, M., Neo-Mahupeleng, G., Ramotadima, M., Rutina, L., Sechele, L. & Sejoe, T.B., Stokke, S., Swenson, J.E., Taolo, C., Vandewalle, M. & Wegge, P. (2004) The return of the giants: Ecological Effects of an Increasing Elephant Population. Ambio, 33(6), 276-282.  Smit, I.P.J. & Ferreira, S.M. (2010) Management intervention affects river-bound spatial dynamics of elephant. Biological Conservation, 143, 2172-2181.  Spanbauer, B.R. & Adler, G.H. (2015) Seed protection through dispersal by African savannah elephants (Loxodonta africana africana) in northern Tanzania. African Journal of Ecology, 53, 496-501.  Staub, C.G., Binford, M.W. & Stevens F.R. (2013) Elephant herbivory in Majete Wildlife Reserve, Malawi. African Journal of Ecology, 51(4), 536-543.  Stockwell, C.A., Hendry, A.P. & Kinnison, M.T. (2003) Contemporary evolution meets conservation biology. Trends in Ecology & Evolution, 18, 94–101.  Stokke, S. & du Toit, J.T. (2000) Sex and size related differences in the dry season feeding patterns of elephants in Chobe National Park, Botswana. Ecography, 23, 70-80.  Stokke, S. & du Toit, J.T. (2002) Sexual segregation in habitat use by elephants in Chobe National Park, Botswana. African Journal of Ecology, 40, 360-371.  Trimble, M.J., Ferreira, S.M. & van Aarde, R.J. (2009) Drivers of megaherbivore demographic fluctuations: inference from elephants. Journal of Zoology, 279, 18-26.  Van Aarde, R.J. & Jackson, T.P. (2007) Megaparks for metapopulations: Addressing the causes of locally high elephant numbers in Southern Africa. Biological Conservation, 134, 289-297.

Stellenbosch University https://scholar.sun.ac.za

 Walsh, P.D., White, L.J.T., Mbina, C., Idiata, D., Mihindou, Y., Maisels, F. & Thibault, M. (2001) Estimates of forest elephant abundance: projecting the relationship between precision and effort. Journal of Applied Ecology, 38, 217–228.  Wasser, S., Poole, J., Lee, P., Lindsay, K. & Dobson, A. (2010) Elephants, Ivory, and Trade. Science, 327, 1331–1332.  Weinand, J.J. (2013) Woody vegetation change and elephant water point use in Majete Wildlife Reserve: implications for water management strategies. MSc thesis, University of Stellenbosch, Stellenbosch, South Africa.  Whitehouse, A.M. & Hall-Martin, A.J. (2000) Elephants in Addo Elephant National Park, South Africa: reconstruction of the population’s history, Oryx, 34(1), 46-55.  Whitehouse, A.M., Hall-Martin, A.J. & Knight, M.H. (2001) A comparison of methods used to count the elephant population of the Addo elephant National Park, South Africa. African Journal of Ecology, 39, 140-145.  Whyte, I.J. & Grobler, D.G. (1998) Elephant Contraception research in the Kruger National Park. Pachyderm, 25, 45-52.  Whyte, I.J., van Aarde, R.J. & Pimm, S.L. (2003) Kruger’s Elephant Population: Its Size and Consequences for Ecosystem Heterogeneity. The Kruger Experience: Ecology and Management of Savanna Heterogeneity. (eds J.T. du Toit, K.H. Rogers & H.C. Biggs), pp 332- 348. Island Press, Washington.  Williamson, B. R. (1976) Reproduction in female African elephants in the Wankie National Park, Rhodesia. South African Journal of Wildlife Research, 6, 89-93.  Wittemyer, G., Daballen, D. & Douglas-Hamilton, I. (2014) Comparative Demography of an At- Risk Elephant Population. Plos One, 8(1), e53726.  Woolley, L-A., Millspaugh, J.J., Woods, R.J., van Rensburg, S.J., Page, B.R. & Slotow, R. (2008) Intraspecific Strategic Responses of African Elephants to Temporal Variation in Forage Quality. Journal of Wildlife Management, 73(6), 827-835.

Stellenbosch University https://scholar.sun.ac.za

Chapter 4

A preliminary investigation of the diet of African elephant (Loxodonta africana) in Majete Wildlife Reserve, Malawi, using stable isotope analysis

F.A. Forrer 1, A. Leslie1

1Department of Conservation Ecology and Entomology, Stellenbosch University, Matieland, Western Cape, 7602, South Africa

Abstract

African elephants (Loxodonta africana) are classified as mixed feeders as they consume a combination of both C3 browse and C4 grass species. The aim of this study was to conduct a preliminary investigation to determine the seasonal dietary components of the African elephant in Majete Wildlife Reserve, Malawi, using stable isotope analysis of faecal samples. Stable isotope analysis of herbivore faecal matter is a method that allows one to determine the proportions of C3 and C4 resources in the diet of herbivore species and so provides a greater understanding of the feeding ecology in general.

Elephants’ diet displayed a clear seasonal difference in the proportion of C3 browse consumed. In the dry seasons their diet consisted of 98% C3 browse. The proportion of C3 vegetation in the diet decreased to 59% in the early wet season and to 65% in the late wet season, indicating a greater proportion of C4 vegetation being consumed in the wetter seasons. The results of this study will contribute towards a greater understanding of the ecological requirements of elephants in Majete Wildlife Reserve and will allow wildlife managers to make informed decisions with regards to population management strategies.

Keywords: African elephant, browser, carbon-13, grazer, stable isotope analysis

Stellenbosch University https://scholar.sun.ac.za

Introduction

The African elephant (Loxodonta africana) is known as a ‘keystone’ species as it plays a pivotal role in structuring both plant and animal communities in the habitats it occupies (Blanc et al. 2007). Elephants often dominate mammal biomass and the total mass of vegetation consumed by elephants on a daily basis, when compared to other herbivore species, is far greater due to their body size (White 1994). It is for this reason that elephants’ are known agents of vegetation change in African woodlands, especially when high densities of the species build up in confined areas (Guy 1976; Cochrane 2003; Western & Maitumo 2004; Guldemond & van Aarde 2007). In Sengwa Wildlife Research Area, Zimbabwe, high elephant densities changed the structure and composition of miombo woodland, causing a 48% decrease of large trees and significantly increasing small trees and shrubs (Mapaure & Moe 2009). Elephants were responsible for local floristic changes as the dominant tree species, Brachystegia boehmii, was replaced by a less palatable species, namely Pseudolachnostylis maprouneifolia (Mapaure & Moe 2009). Therefore, due to their body size and the amount of vegetation consumed, elephants act as ecological engineers although this varies according to population densities and the ability of a population to migrate (Guldemond & van Aarde 2007; Loarie et al. 2009a).

In the past, elephant phylogenetic lineages have displayed both browsing and grazing members (Cerling et al. 1999). The elephant lineages, Elephas and Loxodonta, have been traced back as far as 5 million years to the late Miocene era of Africa where their initial appearance coincides with the worldwide expansion of C4 biomass (Cerling et al. 1999). The ancestors of modern day elephants were almost entirely C4 grazers until about 1 million years ago when they started to consume more C3 browse species (Cerling et al. 1999). Modern Elephas and Loxodonta are almost exclusively browsers in all environments they occur in and so it is a possibility that they are in another evolutionary transition from being predominantly grazers to being predominantly browsers (van der Merwe et al. 1988; Cerling et al. 1999). Evidence of this is displayed in the fact that the majority of elephant populations have mixed C3/C4 diets whilst only a few select populations, such as in Addo Elephant National Park (hereafter, AENP), South Africa, and Tsavo (hereafter, TENP), Kenya, display near pure

C4 diets (van der Merwe et al. 1988; Cerling et al. 1999). Concurrently, extant elephant molar teeth have retained a high crowned form, which has long been the dental interpretation of a modification for abrasive grazing diet, yet in the present day the majority of elephant populations have large proportions of browse in their diet (Cerling et al. 1999; Cerling et al. 2004). Therefore, this further supports the hypothesis that modern day L. africana are transitioning to a C3 browse based diet rather than maintaining a C4 grazing diet.

Stellenbosch University https://scholar.sun.ac.za

Despite the change in the African elephants’ diet, the fact that they consume a combination of both

C3 and C4 plant species classifies them as mixed feeders (van der Merwe et al. 1988; Cerling et al. 1999; Cerling et al. 2004; Codron et al. 2006). Elephants are selective feeders and tend to prefer plant species such as, for example Acacia, Colophospermum and Adansonia (Cerling et al. 2004; Staub et al. 2013). The ratio of browse and grass in an elephants diet, as well as their preferred plant species, varies widely and therefore contention persists within African elephant literature regarding whether elephants are primarily browsers or grazers (Codron et al. 2006). For example, in the Bunyoro region in Uganda grazing was found to be a priority for elephants (Wing & Buss 1970). Whereas other observers have noted an important browsing component in elephants diet (Jackman & Bell 1884; Codron et al. 2006). Elephant diet preferences, therefore, depend on a number of factors including the region, water availability, rainfall, elevation, soil nutrient composition and vegetation cover (van der Merwe et al. 1988; Cerling et al. 2004; Codron et al. 2006; Loarie et al. 2009b).

Stable isotope ecology provides a means to quantify the consumption of browse and graze by various herbivore species (Cerling, et al. 1999; Codron et al. 2006). The basics of stable isotope ecology are the ratios of carbon (δ13C) and nitrogen (δ15N) as these two elements both appear in a consumer’s tissues and their values mimic the isotopic signatures of the food that was ingested (Codron et al. 2007). In subtropical savannas the bimodal distribution of δ13C between plants photosynthetic pathways (C3 and C4) is recorded in the tissues of the consumer (Codron et al. 2007). Stable isotope ratios are distinct between plants; trees, shrubs and forbs tend to have C3 photosynthetic pathways

13 and so δ C values range between -22 and -35%o and average around -27 ‰, whereas grasses tend to

13 have C4 photosynthetic pathways and δ C values that range between -10 and -15‰ (Smith & Epstein 1971; Cerling et al. 1999). Additionally, animal δ15N values reflects the trophic position, it increases on different levels of the food chain and varies with the rainfall, ecophysiology, protein uptake and nutritional stress (Codron et al. 2007). Thus, stable isotope ecology provides a method of determining the quantity of carbon and nitrogen in the consumer’s tissues, essentially allowing the proportion of graze and browse and the trophic levels to be determined (Cerling et al. 1999; Cerling et al. 2004; Codron et al. 2006; Codron et al. 2007).

When conducting stable isotope analysis, the different materials analysed allow the study of the diet of species at different time scales. Long lived tissues such as hair and teeth integrate an average over months or years and when sampled in series reveal dietary variations within the individual at this scale (Cerling et al. 1999; Cerling et al. 2004; Cerling et al. 2009). Animal faeces, on the other hand, is a useful indicator of dietary variation as it shows high resolution time scales, in order of several days rather than months or years (Codron et al. 2005; Codron et al. 2006; Codron et al. 2007; Codron et al. 2011). Due to this rapid turnover of data no mathematical correction is necessary as the differential

Stellenbosch University https://scholar.sun.ac.za

growth rate and attenuation time is significantly less compared to the use of hair and teeth (Codron et al. 2007). Additionally, faeces are easy to sample in the field and sampling does not require interference through the manipulation or slaughter of animals (Codron et al. 2005; Codron et al. 2006; Codron et al. 2007). There are a number of studies that have shown the relationship and consistency

13 15 between δ C, less so δ N in C3 and C4 diets and faeces, and despite the fact that faeces represents the undigested portion of the diet, it is a relatively accurate reflection of the true diet of the individual (Codron et al. 2005; Codron et al. 2006; Codron et al. 2007).

Defining and understanding the extent of dietary flexibility of herbivores is important to the development and implementation of suitable management practises in conservation. This preliminary investigation aimed to determine the seasonal dietary components of the African elephants in Majete Wildlife Reserve, Malawi, using stable isotope analysis of faecal samples. The analyses were performed on faeces as it reflects the diet of the animal 2 to 3 days before deposition, allowing the investigation of seasonal change as the time of deposition will accurately reflect the time of year (Sponheimer et al. 2003). The fundamental hypothesis was that elephants would have a mixed diet of

C3 browse and C4 graze and that the reflections of these components will vary between the dry season, which will contain more browse, and the wet season, which will contain more graze (Cerling et al. 2004; Codron et al. 2005; Codron et al. 2006; Codron et al. 2007; Cerling et al. 2009). The results and understanding gained will be considered when presenting conservation recommendations to reserve management.

Methods

Study Area Majete Wildlife Reserve (MWR, hereafter) is located in the lower Shire valley region of southern Malawi (S15o 54’26.6”; E034o44’24.3”). It is a 700km2 reserve with northern and eastern boundaries boarded by two perennial rivers, the Mkhulumadzi and the Shire. The altitudinal gradient varies, it is highest in the western region where steeply undulating hills are distorted by river valleys. The altitudinal gradient gradually decreases in the eastern region of the reserve where the terrain flattens towards the Shire River. Majete has two distinct seasons, the wet season that occurs between December and May and the dry season between June and November. The annual precipitation in the reserve is between 680-800mm in the eastern lowlands and 700-1000mm in the western highlands (Wienand 2013). Water availability is affected by season but there are approximately five perennial springs, along with ten artificial waterholes in the reserve (Weinand 2013).

Stellenbosch University https://scholar.sun.ac.za

The vegetation in the reserve is primarily woodland which varies according to altitude. Vegetation types include: high altitude miombo woodland, medium altitude mixed woodland, low altitude mixed woodland and savanna. High altitude miombo woodland occurs in the western hilly region, between 410 – 770m. Species that occur in this vegetation type include Brachystegia boehmii, Julbernardia globiflora and Burkea africana (Patel, H. Unpub). Medium altitude mixed woodland between 230 – 410m, is the ecotonal vegetation that occurs between the western hilly region and the eastern lowlands in the reserve. Vegetation species that occur in this ecotonal region include Brachystegia boehmii, Pterocarpus rotundifolius and Combretum species (Patel, H. Unpub). The low altitude mixed woodland in the eastern region occurs between 205 – 280m. Species that occur in this vegetation type include Acacia species and Sterculia species (Patel, H. Unpub). Lastly, savanna occurs in patches throughout MWR with a slightly higher concentration in the south east of the reserve, with Combretum and Panicum being the dominant species (Patel, H. Unpub)

MWR was gazetted in 1955 but by 2003 most of its large game had been decimated due to poaching and poor management. From 2003 - 2010, MWR underwent one of Africa’s greatest reintroduction programmes, after a Public Private Partnership (PPP) agreement was made between African Parks, Majete (Pty) Ltd. and the Malawian government in early 2003. Wildlife reintroductions to MWR were undertaken in stages as the reserve was not yet fenced. The first stage involved fencing a smaller area of 140km2 in the north-eastern region of the reserve in 2003, known as the Sanctuary, purely for reintroduction purposes. The second stage fenced the remaining boundary of the reserve which was completed in 2008, after which the initial sanctuary fence line was removed in November 2011 and wildlife was able to roam into the greater reserve. In total over 2550 individual animals of 14 different species were reintroduced to MWR. Species reintroduced into the reserve included elephant, black rhino, buffalo, sable hartebeest and a number of other antelope species.

Faecal and Vegetation Sampling and Analysis To determine seasonal changes in the proportion of graze and browse for elephant in MWR, representative vegetation samples and elephant faecal matter samples were collected, identified and dried over a period of 14 months. Both the faeces and the vegetation samples were used for the stable isotopic analysis.

Elephant faecal samples were collected on a weekly basis throughout MWR from March 2015 to the end of February 2016, so as to compare the grass and browse proportions for both the wet and dry seasons. The seasons were defined as follows: late wet season (March – May 2015), early dry season (June – August 2015), late dry season (September – November 2015) and early wet season (December 2015 – February 2016). Only recently deposited, fresh or damp, faeces (not older than 12 hours) were

Stellenbosch University https://scholar.sun.ac.za

collected to ensure that samples represented the appropriate season and were not contaminated by fungi, soil and insects (Wrench et al. 1996). Samples, the size of a small handful, were collected from the centre of two to three different faecal boluses of the same faecal pile and placed into brown envelopes. Each separate faecal pile encountered was taken to represent a different individual of the relevant species. All faecal samples were immediately left to dry in the sun until all moisture was evaporated and samples were thoroughly dry. Data such as the date, season, GPS location and species were recorded when samples were collected. A total of 67 faecal samples were collected, however due to mould only 40 samples were isotopically analysed.

To better interpret the carbon isotopic results from the faecal samples, local vegetation samples were also collected. Vegetation samples were taken from various tree, forb and grass species from within twelve circular transects (approximately 10m in diameter) established in each of the different habitats (woodland below 250m, woodland between 250 – 400m, woodland above 400m and savanna) throughout MWR (described in detail by Codron et al. 2005). Only a small sample, a few stems and leaves, of each plant species was collected and placed in brown paper envelopes for drying purposes. The date, species, GPS location and habitat type at the time of collection was recorded. A total of 140 plant samples were collected, of which 14 different tree species, 5 shrub species and 5 grass species were isotopically analysed. Both faecal and plant sampling was concentrated around the north-east (the Sanctuary) and the south-east region (Pende) of MWR. Additional data on δ13C and δ15N values of grasses and shrubs was acquired from Spies (2015).

All faecal and vegetation samples were oven dried at 60oC for 24 hours and homogenized into a powder using a Beadbug Microtube Homogeniser 115V (USA), set at a speed of 330rmp for 3 minutes. Each sample was processed 3 times in the homogeniser in order to obtain a fine powder that was able to pass through a 1mm sieve. Aliquots of approximately 1.0 to 1.1 mg (faeces & plant material) were weighed into tin capsules (RJM Systems (Pty) Ltd., Product number D1006, Tin Capsules Pressed, Standard Weight 6x4mm) and placed in assay trays ready for analysis. Carbon isotope ratios and percent nitrogen were obtained by placing each weighed sample separately into a Flash EA 1112 Series coupled to a Delta V Plus stable light isotope ratio mass spectrometer via a ConFlo IV system (all equipment supplied by Thermo Fischer, Bremen, Germany), housed at the Stable Isotope Laboratory, Mammal Research Institute, University of Pretoria. A laboratory running standard (Merck Gel: δ13C = -20.57‰, δ15N=6.8‰, C%=43.83, N%=14.64) and blank sample were run after every 12 unknown samples.

Stellenbosch University https://scholar.sun.ac.za

13C/12C ratios are expressed in the standard delta (δ) notation in per thousand (‰) relative to Vienna Pee-Dee Belemnite (VPDB) standard and to air for the nitrogen isotopes values (Coplan 1996; Radloff et al. 2013). Results are expressed using the standard equation:

δX(‰) = [(Rsample-Rstandard)/Rstandard-1]x1000

Where X = 15N or 13C and R represents 15N/14N or 13C/12C respectively. Analytical precision was <0.05‰ for δ13C and <0.13‰ for δ15N.

Statistical Analysis 13 To determine whether there was a significant difference between the δ C values of C3 and C4 plants a two tailed t-test was used (Statistica 13). Additionally, to determine whether there was a significant difference between elephant faeces δ13C values between seasons, a one-way analysis of variance

(ANOVA) was used (Statistica 13). To determine the proportional contribution of C3 plants (browsing) in each season the Bayesian stable-isotope mixing model SIAR in R 3.3.1 was used (Phillips 2001;

13 Phillips & Gregg 2001; Phillips & Koch 2002; Parnell et al. 2010). The average δ C values for C3 and C4 plants was determined as -29.11‰ and -12.92‰ respectively and were used as endpoint values.

13 Discrimination factors (δ C) from plants to faeces of +1.1 ±0.3‰ for C4 plants, and +0.7 ±0.3‰ for C3 plants were entered into the system and the concentration was set to zero (Sponheimer et al. 2003; Codron et al. 2005; Radloff 2013). Median and 95% credibility intervals (Cr.I) were determined and all isotope values are presented as mean ±SD. However, when not working in R a diet-faeces fractionation of -0.9% was used (Sponheimer et al. 2003; Codron et al. 2005).

Results

13 The plants in MWR had δ C values entirely consistent with expectations for C3 and C4 vegetation.

13 15 13 15 Uncorrected δ C and δ N values as well as the standard corrected δ C and δ N values for the C3 and

13 C4 vegetation samples are presented in Table 4.1. There was a notable difference between the δ C values of C3 compared to C4 plants as the trees and forbs were significantly depleted in carbon compared to the grasses (Mann-Whitney U test; z-value=3.85, p=0.000120). The standard corrected

13 δ C values for the 19 C4 samples had a range from -11.47‰ to -14.44‰ and an average of -12.92‰

13 (±0.96). Whereas the standard corrected δ C values of the 34 C3 samples had a range of -26.45 ‰ to -31.54‰ and an average of -29.11‰ (±1.26).

15 Additionally, the δ N values represent the crude protein content of each plant sample. For C4 plants the δ15N values ranged from -2.53% to 7.72% with an average of 0.91% (±2.76) and had a broader

15 range and a greater average value than the nitrogen δ N values for the C3 plants. These values ranged

Stellenbosch University https://scholar.sun.ac.za

from -3.55% to 3.86% with an average of 0.03% (±1.88), however the difference between the δ15N values for C3 and C4 plants was not statistically significant (Mann-Whitney U test; z-value=0.58, p=0.561276).

13 15 Table 4.4. δ C and δ N values (‰) of C3 shrubs and trees and C4 grass specimens used as a reference in the stable isotope analysis of the diet of elephant in MWR, Malawi. The standard corrected values were determined by the Isotope Laboratory, Mammal Research Institute, University of Pretoria.

δ15N (Std δ13C (Std Species Type C3/C4 δ15 N δ13 C C:N Ratio corrected) corrected)

Heteropogon contortus Grass C4 -1.55 -1.60 -14.73 -13.55 43.18 Heteropogon contortus Grass C4 -1.78 -1.83 -14.60 -13.44 47.78 Heteropogon contortus Grass C4 -1.26 -0.80 -12.17 -12.17 21.75 Heteropogon contortus Grass C4 0.86 1.30 -11.70 -11.72 25.31 Heteropogon contortus Grass C4 -0.85 -0.39 -11.56 -11.58 32.75 Erogrostis aspera Grass C4 -0.86 -0.90 -15.12 -13.93 45.04 Urochloa mosambiscensis Grass C4 1.15 1.10 -14.42 -13.23 39.85 Urochloa mosambiscensis Grass C4 0.82 1.27 -12.48 -12.47 19.77 Urochloa mosambiscensis Grass C4 0.73 1.17 -12.57 -12.56 16.25 Panicum maximum Grass C4 -0.83 -0.88 -14.79 -13.62 45.22 Cynodon dactylon Grass C4 3.16 3.58 -14.32 -14.27 18.33 Cynodon dactylon Grass C4 2.30 2.73 -13.85 -13.80 16.85 Cynodon dactylon Grass C4 2.49 2.92 -14.50 -14.44 24.78 Dactylotenium aegyptium Grass C4 1.66 2.10 -13.72 -13.68 24.24 Dactylotenium aegyptium Grass C4 7.34 7.72 -12.87 -12.85 17.31 Dactylotenium aegyptium Grass C4 5.37 5.77 -13.12 -13.09 26.53 Hyparrhenia rufa Grass C4 -1.97 -1.50 -11.60 -11.62 34.01 Hyparrhenia rufa Grass C4 -3.01 -2.53 -11.93 -11.94 40.68 Hyparrhenia rufa Grass C4 -2.40 -1.93 -11.45 -11.47 29.09 Neorautanenia mitis Shrub C3 -1.90 -1.95 -30.63 -29.47 14.35 Becium grandiflorum Shrub C3 0.24 0.19 -30.95 -29.79 22.08 Pilea tetraphylla Shrub C3 0.09 0.04 -32.70 -31.51 14.42 Asystasia gangetica Shrub C3 2.25 2.50 -30.07 -28.88 16.82 Grewica Bicolour Shrub C3 0.44 0.39 -30.71 -29.54 22.33 Brachystegia utilis Tree C3 -0.17 -0.22 -27.63 -26.45 35.00 Monotes africanus Tree C3 -0.37 -0.24 -30.79 -29.61 29.78 Diospyros kirkii Tree C3 -2.92 -2.96 -30.15 -28.97 36.49 Diospyros kirkii Tree C3 -2.92 -3.03 -30.16 -29.02 37.15 Julbernardia globiflora Tree C3 0.47 0.81 -28.76 -27.60 25.59 Julbernardia globiflora Tree C3 0.86 0.42 -28.61 -27.47 26.02 Pseudolachnostylis maprouneifolia Tree C3 -3.09 -3.13 -30.05 -28.90 30.89 Pseudolachnostylis maprouneifolia Tree C3 -0.37 -0.42 -29.17 -27.97 26.04 Acacia nilotica Tree C3 -1.27 -1.15 -31.63 -30.45 30.18 Combretum adenogonium Tree C3 -3.15 -3.55 -29.54 -28.36 35.18 Sterculia quinqueloba Tree C3 2.23 2.18 -30.92 -29.74 31.16 Terminalia Sericea Tree C3 -1.41 -1.46 -29.80 -28.63 38.38

Stellenbosch University https://scholar.sun.ac.za

Diplorhynchus condylocarpon Tree C3 -0.17 -0.22 -29.95 -28.78 26.64 Diplorhynchus condylocarpon Tree C3 -1.28 -1.33 -29.49 -28.34 29.38 Diplorhynchus condylocarpon Tree C3 0.61 1.06 -28.13 -27.72 21.84 Diplorhynchus condylocarpon Tree C3 2.37 2.80 -27.10 -26.72 16.76 Diplorhynchus condylocarpon Tree C3 1.24 1.68 -27.47 -27.08 19.45 Petrocarpus lucens Tree C3 -0.81 -0.86 -30.11 -28.97 18.01 Petrocarpus lucens Tree C3 -0.82 -0.69 -30.25 -29.07 18.42 Acacia nigrescens Tree C3 1.41 1.36 -30.06 -28.90 22.41 Acacia nigrescens Tree C3 3.26 3.68 -30.68 -30.20 11.85 Acacia nigrescens Tree C3 2.51 2.94 -30.00 -29.54 15.98 Acacia nigrescens Tree C3 3.44 3.86 -32.06 -31.54 18.06 Tamarindus indica Tree C3 -1.04 -1.24 -32.14 -30.65 22.38 Tamarindus indica Tree C3 -1.19 -1.08 -31.80 -30.98 20.99 Combretum Zeyheri Tree C3 0.23 0.35 -31.03 -29.85 23.16 Dalbergia melanoxylon Tree C3 -0.35 0.10 -29.19 -28.75 12.61 Dalbergia melanoxylon Tree C3 -0.64 -0.18 -30.35 -29.78 12.38 Dalbergia melanoxylon Tree C3 0.03 0.48 -31.00 -30.51 14.12

The uncorrected δ13C and δ15N values as well as the standard corrected δ13C and δ15N values for the elephant faecal samples collected in the early dry, early wet, late dry and late wet seasons are presented in Table 4.2. The δ13C values for the faeces ranged from -18.29‰ to -30.49‰ over the entire sampling period. In the early dry season δ13C values ranged from -27.30‰ to -30.49‰ with an average of -29.16‰ (±0.82). In the early wet season δ13C values ranged from -18.29‰ to -29.12‰ with an average of -22.22‰ (±4.17). In the late dry season δ13C values ranged from -28.04‰ to - 29.26‰ with an average of -28.65‰ (±0.39). In the late wet season δ13C values ranged from -19.48‰ to -26.27‰ with an average of -23.01‰ (±2.52).

A Kruskal-Wallis test revealed significant differences in faecal carbon (δ13C) between seasons; the early dry season and the early wet season (z-value=4.29, p=0.000105) and the early dry season and the late wet season (z-value=4.29, p=0.000107). There was also a significant difference in faecal carbon (δ13C) between the late dry season and the early wet season (z-value=3.05, p=0.013879) and the late dry season and the late wet season (z-value=2.98, p=0.017070). However, there was no significant difference in faecal carbon (δ13C) between the early and late dry seasons (z-value=1.27, p=1.00) and the early and late wet seasons (z-value=0.20, p=1.00). Additionally, the nitrogen δ15N values for the elephant faeces ranged from -0.33 % to 1.88% over the entire sampling period with an average of 0.71% (±0.59).

Stellenbosch University https://scholar.sun.ac.za

Table 4.5. δ15N and δ13C values (‰) of faecal samples representing the diet of elephant in the early dry season 2015, late dry season 2015 an the late wet season 2015 and early wet season 2016 in MWR, Malawi. The standard corrected values were determined by laboratory technicians at the Stable Isotope Lab, Mammal Research Institute, University of Pretoria.

δ15N (Std δ13C (Std Species Season δ15 N δ13 C C:N Ratio corrected) corrected)

Elephant Early Dry 1.51 1.62 -30.49 -29.31 43.90 Elephant Early Dry 0.50 0.60 -29.88 -28.69 45.93 Elephant Early Dry 0.56 0.67 -30.53 -29.34 45.28 Elephant Early Dry -0.02 0.10 -30.34 -29.15 51.27 Elephant Early Dry 1.41 1.52 -30.97 -29.79 51.20 Elephant Early Dry -0.15 -0.03 -28.48 -27.30 48.72 Elephant Early Dry 1.02 1.14 -31.67 -30.49 31.11 Elephant Early Dry -0.10 0.01 -29.64 -28.46 47.99 Elephant Early Dry 0.41 0.52 -30.39 -29.20 56.99 Elephant Early Dry 0.64 0.75 -30.93 -29.75 51.14 Elephant Early Dry 0.75 0.86 -30.46 -29.28 52.03 Elephant Early Wet 1.79 1.88 -22.64 -21.44 32.16 Elephant Early Wet 0.85 0.94 -20.01 -18.82 42.81 Elephant Early Wet -0.12 -0.04 -30.32 -29.12 51.44 Elephant Early Wet 0.27 0.35 -22.65 -21.45 42.25 Elephant Early Wet 1.61 1.69 -19.79 -18.59 31.63 Elephant Early Wet -0.20 -0.12 -21.03 -19.83 32.25 Elephant Early Wet 0.51 0.59 -19.49 -18.29 35.80 Elephant Early Wet 0.51 0.59 -27.19 -26.00 28.71 Elephant Early Wet 0.96 1.04 -29.94 -28.74 54.90 Elephant Early Wet 1.42 1.50 -21.15 -19.96 34.65 Elephant Late Dry 0.52 0.62 -30.14 -28.95 44.28 Elephant Late Dry 0.92 1.01 -29.86 -28.67 40.22 Elephant Late Dry 0.02 0.12 -29.66 -28.47 54.87 Elephant Late Dry -0.24 -0.14 -30.44 -29.26 46.93 Elephant Late Dry -0.43 -0.33 -30.45 -29.26 38.84 Elephant Late Dry -0.20 -0.10 -29.23 -28.04 43.62 Elephant Late Dry 1.33 1.43 -29.89 -28.70 45.27 Elephant Late Dry 0.70 0.79 -29.57 -28.38 46.32 Elephant Late Dry 0.49 0.58 -29.61 -28.42 46.96 Elephant Late Dry 0.28 0.38 -29.44 -28.26 53.26 Elephant Late Dry -0.31 -0.21 -30.00 -28.81 41.64 Elephant Late Wet 1.00 1.06 -22.43 -21.23 29.00 Elephant Late Wet 0.27 0.34 -22.66 -21.46 38.15 Elephant Late Wet 1.41 1.48 -23.03 -21.83 30.87 Elephant Late Wet 0.69 0.76 -22.67 -21.47 33.23 Elephant Late Wet 1.48 1.55 -27.87 -26.67 27.23 Elephant Late Wet -0.05 0.02 -22.93 -21.73 44.06 Elephant Late Wet 0.97 1.05 -27.11 -25.91 28.71 Elephant Late Wet 1.03 1.10 -25.90 -24.70 32.13 Elephant Late Wet 0.55 0.61 -27.42 -26.21 33.83 Elephant Late Wet 0.56 0.63 -26.29 -25.09 32.59

Stellenbosch University https://scholar.sun.ac.za

Elephant Late Wet 0.66 0.72 -20.68 -19.48 30.69 Elephant Late Wet 1.41 1.47 -21.58 -20.37 35.50

15 13 The δ N and δ C values of C3 and C4 vegetation as well as the elephant faecal samples collected in the various seasons is displayed in Figure 4.1. There is a clear, almost predominant contribution of C3 vegetation to the diet of elephant in the early dry and late dry season. Whereas, in the early and late wet seasons elephants displayed a dietary shift towards the C4 vegetation, indicating a greater diversity of browse (C3) and grass (C4) in the diet compared to the early and late dry seasons. The contribution of C3 browse to the diet of elephants ranged from 59% to 98% for the entire sampling period (Table 4.3.). More specifically, in the early dry and late dry season the proportion of C3 in the diet of elephant averaged 98% (95% Cr.I.: 96-100%) and 98% (Cr.I.: 95-100%), indicating that the diet consisted almost exclusively of C3 browse vegetation. However, the proportion of C3 vegetation in the diet decreased in the early wet season to an average of 59% (Cr.I.: 45-75%) and in the late wet season to 65% (Cr.I.: 56-74%), indicating a greater proportion of C4 vegetation being consumed in the wet seasons.

C4 3 C3 15

δ N Early Dry 1 Early Wet

-1 Late Dry Late Wet

-3

-5 -35 -30 -25 -20 -15 -10 δ C13

Stellenbosch University https://scholar.sun.ac.za

Table 4.6. Percentage browse consumption of elephant in MWR, Malawi, as determined from stable carbon isotope analysis of faecal samples (n=number of samples). Percentage browse consumption was determined using a Bayesian stable-isotope mixing model using C3 (-29.23‰) and C4 (-13.12‰) plant end-members.

Elephant Elephant % browse Season n δ13C (‰) ±SD (95% Cr.I) Early Dry 11 -29.16 0.82 98 (96-100) Early Wet 10 -22.22 4.17 59 (45-75) Late Dry 11 -28.65 0.39 98 (95-100) Late Wet 12 -23.01 2.52 65 (56-74)

Discussion

The main objective of this study was to investigate whether season had an effect on the diet of elephants in MWR, Malawi. In order to achieve this objective only plant δ13C values were corrected with discrimination factors in the SIAR package (R 3.3.1) and analysed to determine the proportion of

15 C3 and C4 biomass in the faecal samples. The δ N values were of little use as the current understanding of nitrogen isotopes is poor and the variation of nitrogen isotopic values in plant and animal faecal matter is more complex than that of the carbon isotopes due to environmental factors affecting nitrogen fixation and fractionation. Researchers have thus been cautioned when analysing and interpreting δ15N values (Sandberg et al. 2012; Dr Frans Radloff per. Comms).

According to a number of studies, the isotopic classification of elephant’s diets is mixed feeding but depending on their location, certain populations may have near-pure grass- or browse-based diets (Wing & Buss 1970; Barnes 1983; Cerling et al. 1999; de Boer et al. 2000; Codron et al. 2005; Cerling et al. 2004; Codron et al. 2006; Codron et al. 2011). In this study, there was a significant change in elephant diet from the early and late dry seasons, where the majority (98%) of food consumed was C3 browse, to the early and late wet seasons where the proportion of browse consumed decreased to

59% (Cr.I:45-75%) and 65% (Cr.I: 56-74%) respectively. This concurs with the prediction that the C4 content of the elephant’s diet would increase in the wet season due to fresh grass growth providing greater palatability, lower tannin concentrations and higher nutritional value compared to many C3 browse species (Cooper & Owen-Smith 1985; Sukumar & Ramesh 1992; de Boer et al. 2000). Therefore, this finding is supported by numerous other studies that suggest elephants have a dietary shift from a nearly pure C3 browse diet in the dry seasons to a greater proportion of C4 plants being consumed in the wetter seasons (Barnes 1983; van der Merwe et al. 1988; Cerling et al. 1999; Codron et al. 2005; Cerling et al. 2004; Codron et al. 2006; Codron et al. 2011).

Additionally, the proportion of C4 graze consumed in this study, 2% in the drier seasons, 41% in the early wet and 35% in the late wet, is comparable to other studies that have used similar methods to

Stellenbosch University https://scholar.sun.ac.za

determine the seasonal proportion of C3 and C4 biomass in elephant diets. Codron et al. (2006) found that elephants in the northern Kruger National Park, South Africa, had a diet composed of 40% C4 graze in the dry season and 50% graze in the wet season, whereas elephants in the southern areas of the park had a diet composed of 10% C4 graze in the dry season and 50% graze in the wet season.

Cerling et al. (2009) found that elephants in northern Kenya had a diet composed of 5-20% C4 graze in the dry season and 40-60% C4 graze in the wet season. Throughout isotopic dietary studies conducted on elephant, very few elephant populations have been recorded as essentially pure C4 grazers in the wet season. This is supported by van der Merwe et al. (1988) who found that elephants from Namibia,

South Africa, Botswana, Zambia and Malawi have an essentially pure C3 diet with very little C4 intake. The author did, however, find an exception in Kenya as the elephant’s diet had a significantly large proportion of C4 graze in the diet. The author credited this to the “elephant problem” where by a high number of elephants reduced the woodlands, that were gradually replaced by extensive grass cover and so elephants were largely unable to access C3 browse (Napier Bax 1963, Cerling et al. 2004).

Therefore, it can be suggested that the proportion of C3 and C4 biomass in elephant diets can vary according to location as well as seasonal rainfall. Although, in this study the elephants of MWR tended to consume far less C4 graze in the drier seasons compared to the other studies, this could be attributed to the severity of the dry season in the Shire basin, where MWR is located, whereby the extreme high temperatures and lack of water diminish the available vegetation and in particular the grass species.

Apart from the data displaying a near pure C3 browse diet in the dry season, the wet season samples displayed a wide range of δ13C values in both the early and late wet seasons. The results suggest that individuals have both a mixed feeding diet and a pure C3 browse diet in the wetter seasons. This was unexpected as it was thought that when grass is available in the wet seasons, the fact that it has a greater palatability, is higher in nutritional value and generally has lower tannin concentrations compared to many browse species, would make grass the preferred food source (Cooper & Owen- Smith 1985; de Boer et al. 2000; Codron et al. 2006). A possible reason for this result could be location, perhaps certain habitat types and certain areas of MWR contain less grass and preferred browse species in the wetter seasons and so could yield a near pure C3 browse diet. However, the majority of the samples were collected in the Sanctuary (north-west of MWR), where water is plentiful in the form of four artificial waterholes and the Shire River and the habitat type varies between woodland below 250m, woodland between 250-400m and savanna. With so little variation in the location of where the faecal samples were collected it is unlikely that location played a significant role in deciding a near pure C3 diet or a mixed feeding diet. However, the elephant home ranges may very well have played a significant role in the proportion of C3/C4 biomass in the diet. Therefore, the difference in diet in the

Stellenbosch University https://scholar.sun.ac.za

wetter seasons could be attributed to two factors. Firstly, it would be beneficial to understand the home ranges of different individuals in the reserve to fully understand the isotopic results of the seasonal diet and secondly individual personal preference, perhaps some individuals prefer browse species even in the wetter seasons.

However, the fact that this study showed no significance in δ13C values or change in the proportion of

C3 browse in samples within seasons (early and late) is interesting. Typically, because plant growth varies throughout the year according to the amount of rainfall, one may expect change in the proportion of C3 browse and C4 graze in the diets. In the dry season (early and late) the proportion of

C3 browse remains the same, 98%. This suggests that there was very little, if any, grass available throughout the dry season. In addition, van der Merwe et al. (1988) found that elephants in Kasungu

National Park, Malawi, consumed between 30-45% C4 graze in the early wet season, 60% C4 graze in the late wet season, approximately 23% C4 graze in the early dry season and 10% C4 graze in the late dry season. This indicates that there should be a greater change in the proportion of C4 graze within seasons, for example between early wet and late wet seasons. The lack of any significant change in the δ13C values within seasons in this study could be due to the methodology. One way of possibly observing a greater variation within seasons would be to collect and analyse a larger number of samples. As elephants are diverse feeders it would have perhaps been more beneficial to analyse one hundred and twenty faecal samples per month as opposed to just ten per season.

For future studies on elephant diet and the diet of other herbivore species in MWR several important steps need to be taken to improve the current study and to provide a greater in-depth analysis on the seasonal change in diet. The steps are as follows:

1) The reserve should be divided into the four main habitat types: high altitude woodland, medium altitude woodland, low altitude woodland and savanna. These four habitat types should then be divided into sampling blocks according to the size of the habitat and waterhole location. Ideally there should be at least three sampling blocks in each habitat type. (J. Codron per comms.)

2) A thorough vegetation analysis, such as plant transects, should be conducted in each sampling block. A minimum of three different grass species and five different browse species should be collected, with three to five samples of each species, in both the wet and dry season. (Codron et al. 2006; J. Codron per comms.)

3) Sub-sampling of different plant parts will need to accompany the vegetation analysis. Browse species will need separate leaves, fruit, roots and bark sampled. Grass species will need seeds, leaves and roots sampled. The different plant species and their parts will need to be

Stellenbosch University https://scholar.sun.ac.za

isotopically analysed separately to get an idea of the isotopic variation in the plant species in MWR. (J. Codron per comms.)

4) In terms of elephant faecal sampling, ten faecal samples need to be collected per sampling block in each habitat type on a monthly basis. Essentially, a total of 120 samples should be collected monthly. All faecal samples should be analysed for stable carbon and nitrogen isotopes. (Codron et al. 2006; J. Codron per comms.)

5) Previous studies suggest that elephant, in particular, take between twelve and twenty two hours to fully digest their food, from consumption to defecation, and they are capable of moving large distances in a day (Laws & Parker 1968; Rees 1982; Clauss et al. 2007; Campos- Arceiz et al. 2008; Stokke & du Toit 2002). It would be difficult to justify that samples collected in a specific location and habitat type would be a fair representation of that site. This is supported by de Boer et al. (2000) who found that species composition was relatively unaffected by the habitat in which the faecal pile was found and so the long digestive process and the opportunistic feeding pattern of elephant masks the species composition of each habitat (Jarman 1971). One way to overcome this issue is to collect faecal samples of other sympatric herbivores in terms of pure browsers (kudu), pure grazers (zebra) and mixed feeders (impala) (Condron et al. 2006). These additional samples will represent the majority of plant species in the different habitat types more accurately than elephant faecal samples as species such as kudu, zebra and impala do not travel the great distances elephants do on a daily basis (Codron et al. 2006). This will provide an accurate representation of each habitat type as well as a robust isotope baseline for elephants in MWR from both the plant and sympatric herbivore perspective. Therefore, comparing the carbon values between elephant faecal samples to other herbivore species would provide an idea as to whether the elephant faecal samples represent a certain habitat type and allow for the calculations of the distance travelled and the time between consumption and defaecation (J. Condron per comms.).

6) Lastly, in addition to isotope analysis of species faecal matter mictohistological techniques should be used to determine the actual dietary composition of each elephant faecal sample (Stewart 1967). This method allows for the identification of the vegetative remains to a group or species level with the use of dietary reference slides of local vegetation samples. Using this method will give researchers an idea of what vegetation types are being targeted more than others at certain times of the year as well as eliminate the uncertainty that sample location provides. Further details of this method are referred to in Stewart (1967); Suter et al. (2004); de Jong et al. (1995); Blanco-Fontao et al. (2010) and Bergstrom (2013).

Stellenbosch University https://scholar.sun.ac.za

To date three other studies that have focused on the feeding habits of elephant in MWR. Staub et al. (2013) found that browsing of elephant on MWR increased with the proximity to water sources and that a strong preference was shown for riparian woodlands and Acacia dominated woodlands. Whereas Brachystegia dominated woodlands were relatively unaffected by elephant browsing (Jachmann & Bell, 1985; Staub et al. 2013). Wienand (2013) agreed with Staub et al. (2013) and further found that the use of water points by elephant was seasonal, in that visitation of water points in the wet season was minimal compared to the dry season when visitation rates increased. Furthermore, the water points with the highest elephant visitation rates in MWR were situated in low altitude woodland and the water points with the lowest visitation rates were those situated in the high altitude Brachystegia dominated woodlands (Weinand 2013). Komoto (2013) agreed with both the above studies and found a strong correlation between tree mortality and heavy elephant browsing in MWR. The study also proposed that the highest elephant impact on woody tree species was in the north- eastern part of the reserve, the Sanctuary, and the lowest impact was in the Brachystegia dominated woodland (Komoto 2013). Therefore, the water points in the Sanctuary of MWR are most used by elephants due to flat, low altitude terrain, favoured vegetation types are prevalent such as Combretum and Acacia species, there is a high availability of perennial water as both mentioned rivers are accessible.

The combination of the results of these three mentioned studies, as well as the results from this study suggest that the feeding habits of elephants in MWR were strongly seasonal. In the wet season, when water was plentiful and nutritious grass was available, the elephants were able to migrate to areas of the reserve largely inaccessible during other times of year. However, in dry season the low altitudinal areas that contain riparian and Acacia dominated woodland were under increasing pressure from elephants especially the areas closer to perennial water points. This is supported by this study as results indicated that the elephants almost exclusively feed on browse in the dry season in MWR. Therefore, as elephant numbers continue to increase in the reserve, the Sanctuary (the main tourist hub) which contains four artificial water points, a perennial spring, the Shire and Mkulumadzi Rivers, and is low in altitude and has the preferred vegetation types will be under increasing pressure to support a growing elephant population.

The increase in the human population in Southern Africa has resulted in the fragmentation and isolation of many of Africa’s wilderness areas. In order to prevent human-wildlife conflict in the remaining wilderness areas an increasing number of reserves are being fenced to not only protect the local people but also the flora and fauna of the reserves. This ultimately results in an artificial ecosystem as many animal species are unable to migrate and so intensive management is required to preserve the biodiversity of the protected areas. The elephant population in MWR needs to be closely

Stellenbosch University https://scholar.sun.ac.za

monitored to prevent overpopulation and to prevent the species out competing other herbivores and disrupting the balance of the ecosystem as a whole.

Acknowledgements

A big thank you to Dr Alison Leslie for supporting me and all the gallivanting I did for this chapter. Thank you to Dr Grant Hall at the Stable Isotope Laboratory, Mammal Research Institute, University of Pretoria for guiding me with the stable isotope analysis and sharing your knowledge and unique skill set. To Dr Frans Radloff, although our meeting was brief, thank you so much for your time, you truly helped me understand isotopes. To the Earthwatch Institute for funding the Animals in Malawi Research Project, including my own research. Lastly, a big thank you to Claire Gordon and all the Earthwatch volunteers for helping me collect the dung and vegetation samples.

Stellenbosch University https://scholar.sun.ac.za

References

 Barnes, R.F.W. (1983) The Elephant Problem in Rusha National Park, Tanzania. Biological Conservation, 26, 127-148.  Bergstrom, B.J. (2013) Would East African savanna rodents inhibit woody encroachment? Evidence from stable isotopes and microhistological analysis of faeces. Journal of Mammology, 94(2), 436-447.  Blanc, J.J., Barnes, R.F.W., Craig, G.C., Dublin, H.T., Thouless, C.R., Douglas-Hamilton, I. & Hart, J.A. (2007) African Elephant Status Report 2007: an update from the African Elephant Database. Occasional Paper Series of the IUCN Species Survival Commission, No 33. IUCN/SSC African Elephant Specialist Group IUCN, Gland, Switzerland. Vi +276 pp.  Blanco-Fontao, B., Fernandez-Gil, A., Obeso, J.R. & Quevedo, M. (2010) Diet and habitat selection in Cantabrian Capercaillie (Tetrao urogallus cantabricuc): ecological differentiation of a rear-edge population. Journal of Ornithology, 151, 269-277.  Campos-Arceiz, A., Larrinaga, A.R., Weerasinghe, U.R., Takatsuki, S., Pastorini, J., Leimgruber, P., Fernando, P. & Santamaria, L. (2008) Behaviour rather than diet mediates seasonal differences in seed dispersal by Asian elephants. Ecology, 89(10), 2684-2691.  Cerling, T.E., Harris, J.M. & Leakey, M.G. (1999). Browsing and Grazing in Elephants: The Isotope record of Modern and Fossil Proboscideans. Oecologia, 120(3), 364–374.  Cerling, T.E., Passey, B.H., Ayliffe, L.K., Cook, C.S., Ehleringer, J.R., Harris, J.M., Dhidha, M.B. & Kasiki, S.M. (2004) Orphans’ tales: seasonal dietary changes in elephants from Tsavo National Park, Kenya. Palaeoecology, 206, 367 – 376.  Cerling, T.E., Wittemyer, G., Ehleringer, J.R., Remlen, C.H. & Douglas-Hamilton, I. (2009) History of animals using Isotopes Records (HAIR): A 6-year dietary history of one family of African Elephants. PNAS, 106(20), 8093 – 8100.  Clauss, M., Schwarm, A., Ortmann, S., Streich, W.J. & Hummel, J. (2007) A case of non-scaling in mammalian physiology? Body size, digestive capacity, food intake, and ingesta passage in mammalian herbivores. Comparative Biochemistry and Physiology, 148, 249-265.  Cochrane, E.P. (2003) The need to be eaten: Balanites wilsoniana with and without elephant seed-dispersal. Journal of Tropical Ecology, 19(5), 579-589.  Codron, D., Codron, J., Lee-Thorp, J.A., Sponheimer, M., De Ruiter, D., Sealy, J., Grant, R. & Fourie, N. (2007) Diets of savanna ungulates from stable carbon isotope composition of faeces. Journal of Zoology, 273(1), 21-29.

Stellenbosch University https://scholar.sun.ac.za

 Codron, D., Codron, J., Sponheimer, M., Lee-Thorp, J.A., Robinson, T., Grant, C.C. & De Ruiter, D.J. (2005) Assessing diet in savanna herbivores using stable carbon isotope ratios of faeces. Koedoe, 48(1), 115-124.  Codron, J., Codron, D., Lee-Thorp, J.A., Sponheimer, M., Kirkman, K., Duffy, K.J. & Sealy, J. (2011) Landscape-scale feeding patterns of African elephants inferred from carbon isotope analysis of faeces. Oecologia, 165, 89-99.  Codron, J., Lee-Thorp, J.A., Sponheimer, M., Codron, D., Grant, R.C. & De Ruiter, D.J. (2006) Elephant (Loxodonta africana) Diets in Kruger National Park, South Africa: Spatial and Landscape Differences. Journal of Mammalogy, 87(1), 27-34.  Cooper, S.M. & Owen-Smith, N. (1985) Condensed tannins deter feeding by browsing ruminants in a South African savanna. Oecologia, 67(1), 142-146.  Coplan, T.B. (1996) New guidelines for reporting stable hydrogen, carbon, and oxygen isotope ratio data. Geochimica et Cosmochimica Acta, 60, 3359-3360.  de Jong, C. B., Gill, R. M. A., Van Wieren, S. E. & Burlton, F. W. E. (1995) Diet selection by roe deer Capreolus capreolus in Kielder Forest in relation to plant cover. Forest Ecology and Management, 79(1), 91-97.  Guldemond, R. & van Aarde, R. (2007) The impact of elephants on plants and their community variables in South Africa’s Maputaland. African Journal of Ecology, 45, 327-335.  Guy, P.R. (1976) The feeding behaviour of elephant (Loxodonta africana) in the Sengwa Area, Rhodesia. South African Journal of Wildlife Research, 6(1), 55-63.  Jachmann, H. & Bell, R.H.V. (1985) Utilization by elephants of the Brachystegia woodlands of the Kasungu National Park, Malawi. African Journal of Ecology, 23, 245-258.  Jarman, P.J. (1971) Diets of Large Mammals in the Woodlands around Lake Kariba, Rhodesia. Oecologia, 8, 157-178.  Komoto, S.C. (2013) Post translocation browsing impact of African elephant on woody tree species in Majete Wildlife Reserve, Malawi. MSc Thesis, Stellenbosch University.  Loarie, S.R., van Aarde, R.J. & Pimm, S.L. (2009a) Fences and artificial water affect African savannah elephant movement patterns. Biological Conservation, 142(12), 3086-3098.  Loarie, S.R.,van Aarde, R.J. & Pimm, S.L. (2009b) Elephant seasonal vegetation preferences across dry and wet savannas. Biological Conservation, 142(12), 3099-3107.  Mapaure, I. & Moe, S.R. (2009) Changes in the structure and composition of miombo woodlands mediated by elephants (Loxodonta africana) and fire over a 26-year period in north-western Zimbabwe. African Journal Ecology, 47, 175-183.

Stellenbosch University https://scholar.sun.ac.za

 Parnell, A.C., Inger, R., Bearhop, S. & Jackson, A.L. (2010) Source partitioning using stable isotopes: coping with too much variation. Plos ONE, 5(3), e9672.  Phillips, D.L. (2001) Mixing models in analyses of diet using multiple stable isotopes: a critique. Oecologia, 127, 166-170.  Phillips, D.L. and Gregg, W. (2001) Uncertainty in source partitioning using stable isotopes. Oecologia, 127, 171-179.  Phillips, D.L. and Koch, P.L. (2002) Incorporating concentration dependence in stable isotope mixing models. Oecologia, 130(1), 114-125.  Radloff, F.G.T, Van der Waal, C. & Bond. A.L. (2013) Extensive browsing by a conventional grazer? Stable carbon isotope analysis reveals extraordinary dietary flexibility among Sange cattle of North Central Namibia. Isotopes in Environmental and Health Studies, 49(3), 318-324.  Rees, P.A. (1982) Gross assimilation efficiency and food passage time in the African elephant. African Journal of Ecology, 20, 193-198.  Sandberg, P.A., Loudon, J.E. & Sponheimer, M. (2012) Stable isotope analysis in primatology: a critical review. American Journal of Primatology, 0, 1-21.  Spies, K.S. (2015) Ecology of impala (Aepyceros melampus) and waterbuck (Kobus ellipsiprymnus) in Majete Wildlife Reserve, Malawi. MSc Thesis, Stellenbosch University.  Sponheimer, M., Robinson, T., Ayliffe, L., Passey, B., Roeder, B., Shipley, L., Lopez, E., Cerling, T., Dearing, D. & Ehleringer, J. (2003) An experimental study of carbon-isotope fractionation between diet, hair and faeces of mammalian herbivores. Canadian Journal of Zoology, 81(5), 871-876.  Staub, C.G., Binford, M.W. & Stevens F.R. (2013) Elephant herbivory in Majete Wildlife Reserve, Malawi. African Journal of Ecology, 51(4), 536-543.  Stewart, D. R. M. (1967) Analysis of plant epidermis in faeces: a technique for studying the food preferences of grazing herbivores. Journal of Applied Ecology, 4(1), 83-111.  Stokke, S. & du Toit, J.T. (2002) Sexual segregation in habitat use by elephants in Chobe National Park, Botswana. African Journal of Ecology, 40, 360-371.  Sukumar, R. & Ramesh, R. (1992) Stable carbon isotope ratios in Asian elephant collagen: implications for dietary studies. Oecologia, 91, 536-539.  Suter, W., Suter, U., Krusi, B. & Schutz, M. (2004) Spatial variation of summer diet of red deer Cervus elaphus in the eastern Swiss Alps. Wildlife Biology, 10(1), 43-50.  van der Merwe, N.J., Lee Thorp, J.A. & Bell, R.H.V. (1988) Carbon isotopes as indicators of elephant diets and African environments. African Journal of Ecology, 26, 163-172.

Stellenbosch University https://scholar.sun.ac.za

 Western, D. & Maitumo, D. (2004) Woodland loss and restoration in a savanna park: a 20-year experiment. African Journal of Ecology, 42, 111-121.  White, L.J.T. (1994) Biomass of Rain Forest Mammals in the Lope Reserve, Gabon. Journal of Animal Ecology. 63(3), 499-512.  Wienand, J.J. (2013) Woody vegetation change and elephant water point use in Majete Wildlife Reserve: implications for water management strategies. MSc Thesis, Stellenbosch University.  Wing, L.D. & Buss, I.O. (1970) Elephants and Forests. Wildlife Monographs, 19, 2-92.  Wrench, J.M., Meissner, H.H., Grant, C.C. & Casey, N.H. (1996) Environmental factors that affect the concentration of P and N in faecal samples collected for the determination of nutritional status. Koedoe, 39(2), 1-6.

Stellenbosch University https://scholar.sun.ac.za

Chapter Five

Research findings, conclusions and management recommendations for African Parks Majete management staff

Overview

It is anticipated that the high elephant population on Majete Wildlife Reserve (MWR, hereafter) post- reintroduction will begin impacting ecosystem processes, woody vegetation cover and other herbivore species as the pressure on resources such as water and acceptable forage increases. This study essentially investigated and provided baseline information on the size, structure and gender ratio of the elephant population, by means of individual identification techniques, since the reintroductions in 2006, 2008 and 2010. In addition, this study focused on seasonal movement and habitat use of the elephant by employing the use of camera traps positioned in grids and at selected water sources, as well as determining the approximate browse and grass contributions to their diet throughout the year. The findings of this research, along with the supporting information from the literature review, will contribute to the understanding of the baseline ecology of elephant in MWR. In addition, it will assist in the development of management strategies used to control elephant population numbers and mitigate vegetation degradation. In this report, the key study findings are summarised and the management implications of these findings are discussed in detail.

Research Findings

Chapter Three: The population structure and habitat use of African elephant

 The most recent aerial count (Nov 2015) conducted in MWR counted a total of 389 individual elephants and by March 2016, the current study was able to formally identify 366 of these individuals, using mark-recapture techniques. Aerial counts tend to under sample areas, more specifically woodland areas, so it is suggested that the actual population size of elephant is larger than 389.  In the absence of immigration, emigration, human-elephant conflict and heavy poaching the Majete elephant population has grown at an estimated annual rate of 13.8% since the completion of reintroductions in 2010.  The structure of the population was found to be as follows: 11.8% infants (<1 year), 23.8% calves (1-4.9 years), 14.5% juveniles (5-9.9 years), 15.0% small adult (10-19.9 years), 26.8%

Stellenbosch University https://scholar.sun.ac.za

medium adult (20-34.9 years) and 8.2% large adult (>35 years). The sex ratio of individuals below and above the age of 10 years was 1:1.  Camera grid data analysis suggests that elephants prefer areas around water sources in both the wet and the dry seasons. The largest number of individuals recorded in both seasons was in the Sanctuary area and the number of individuals recorded in the Pwadzi and Namisempha regions increased slightly in the dry season.  Water points were used more in the dry season compared to the wet season.  Water points were not equally used in the dry or the wet seasons but least favourable water points, such as Diwa which is situated at a high altitude and is surround by a Brachystegia- dominated vegetation type, experienced an increase in usage since the study conducted by Weinand (2013).  In general the water points with the highest usage were Pende 2 and Pwadzi (based in the low altitudinal region south-west in MWR), the Sanctuary water points had a much lower usage and this is possibly due to the over saturation of perennial water points in the region.  Family herds dominated water point usage in both the wet and the dry season, except in the wet season 2016 where bulls utilised the Sanctuary water points more than family herds. The domination of water points by family herds is possibly due to the constraints of group living as it limits home ranges to perennial water points especially in the dry season, whereas solitary bulls are less water dependent and are able to migrate through large areas of MWR in both seasons.

Chapter Four: Stable isotope analysis of African elephant diet

 The stable isotope analysis of vegetation samples from MWR provided distinct isotopic values

13 of carbon (δ C) for seven grass (C4) species and twenty browse (C3) species. These values were used as reference points for the stable isotope analysis of elephant faeces.  Elephants are mixed feeders that are able to adjust their diet depending on the season. It was

determined that the elephants’ diet had the highest C3 vegetation content in the early and late dry seasons. However, in the early and late wet seasons elephant diet displayed a shift

towards C4 vegetation, signifying a greater diversity of browse (C3) and grass (C4) in the diet compared to the early and late dry seasons. Therefore, elephants consume a higher

proportion of C3 browse relative to C4 grass in the dry seasons and shift their diets in the wetter seasons to incorporate more grass species.

Stellenbosch University https://scholar.sun.ac.za

Conclusions and Management Recommendations

African elephants form an important part of African ecosystems as they play a keystone role in the diversity of habitats they occupy, from deserts and savannas to rainforests (Styles & Skinner 2000). They have an effect on the canopy cover, the distribution of other species and the dispersal of seeds (Styles & Skinner 2000; Shoshani et al. 2004; Kerley & Landman 2006). However, preserving populations of elephants while maintaining biodiversity is a constant challenge for the management of protected areas and more often than not many of these areas are faced with the “elephant problem”. The “elephant problem” is defined by high elephant numbers in conservation areas as they are perceived to alter the vegetation in the area (Caughley 1976; Barnes 1983; van Aarde & Jackson 2007). In the past, the changes in the abundance of elephants, or the alteration of vegetation was viewed as an undesirable disruption of environmental equilibrium conditions and an irreversible threat to biodiversity (Caughley 1976; Barnes 1983; Gillson & Lindsay 2003). However, this concept of the balance of nature is out dated as vegetation composition is now understood to change over time in response to a variety of variables, such as climatic variation, and so there has been a fundamental shift to the flux of nature (non-equilibrium) paradigm, where the ecological processes that promote biodiversity are essentially about change and not stable equilibria (Gillson & Lindsay 2003, Owen- Smith et al. 2006). Therefore, managing elephants at defined management levels is not sensible; numbers should fluctuate through time and space due to the availability of resources, density dependence and environmental forces (Gillson & Lindsay 2003; Skarpe et al. 2004; Owen-Smith et al. 2006).

Despite this shift to the flux of nature paradigm many protected areas are still faced with the problem of high elephant numbers. These high elephant numbers are induced by anthropogenic factors and include the supplementation of water, fencing and the reduction and fragmentation of landscapes that prevent natural migration patterns (van Aarde & Jackson 2007). Artificial water sources allow more extensive dry season ranging, creating the opportunity to overexploit vegetation areas that would have otherwise been inaccessible to elephant except in the wet season (Loarie et al. 2009; Owen-Smith et al. 2006; Kalwiji et al. 2010). Additionally, fences prevent seasonal migration of individuals in and out of an area and so cause elephants to ‘bunch up’ against them during the wet season as their home ranges increase, again locally increasing the pressure of elephants on their resources (Owen-Smith et al. 2006; Guldemond & van Aarde 2007; Loarie et al. 2009). Therefore, the combination of artificial waterholes and fencing creates elephant populations that are most likely to damage the vegetation in an area and the other species that depend on the vegetation, especially during unusually dry years, as elephants are unable to migrate to find food (Shrader et al. 2010).

Stellenbosch University https://scholar.sun.ac.za

Understanding the non-equilibrium paradigm is essential for the management of the elephant population in MWR. The ecological capacity for elephant in a given area is not a fixed figure, but one that varies from year to year and from season to season. Two of the major challenges that face the ecological capacity for elephant are the temporal and spatial scale of the reserve, such as available habitat and rainfall (van Aarde & Jackson 2007; Bothma & du Toit 2010). However, as a broad guideline the ecological capacity for MWR was calculated using two methods. The first is from Bothma & du Toit (2010) which is based on the following equation that was developed for areas where elephants occur naturally and takes into account the reserve size and annual rainfall.

2 Log10large herbivore biomass in kg per km = 1.685 x log10 mean annual rainfall in mm – 1.095

MWR is a 70 000 ha reserve with an annual precipitation of 680 – 800mm in the Eastern lowlands and 700 – 1000mm in the Western Highlands. Rainfall is incredibly important in setting the stocking density of an area and in order to be conservative in the calculation a value of 700mm of annual rainfall was used. Therefore, according to Bothma & du Toit (2010) the recommended ecological capacity of MWR is 304 individual elephants. It must be reiterated, however, that this value is a broad guideline and will vary according to the various habitat types, water supply and other herbivore species and their population sizes.

The second method used to calculate the ecological capacity for elephant on MWR is from van Oudtshoorn (2015) and focuses on the browsing rather than the grazing capacity of elephants. As the results from Chapter 4 indicate that elephants in MWR do not consume significant amounts of grass. The Browsing Units per Hectare (BU/ha) for MWR were calculated using data from Staub et al. (2013) for trees greater than 2.0 m in height, or a Diameter and Breast Height (DBH) greater than 10 cm, and the result was 1937.66 BU/ha. The browsing capacity for elephant for MWR was calculated by dividing the Browsing Units needed by elephant, 41500 BU/needed, by the Browsing Units available, 1937.66 BU/ha, and the result was 21.4 ha/elephant which is equivalent to 402 individual elephants in MWR. Although this value for the browsing capacity of elephants in MWR is suitable for 2013, when Staub et al. (2013) conducted their study, when compared to the current stocking rate of elephants in 2015, which is 22.1 ha/elephant, it is clear that elephants are approaching or have started to exceed the browsing capacity on the reserve. However, these values have only taken elephants into account and other browsers on the reserve, such as Kudu for example, need to be included in the calculations, no doubt changing the browsing capacity for MWR and decreasing the stocking rate of elephant. However, further research is needed regarding the other browsing species and the current state of the woody vegetation in MWR before final numbers for elephant can be provided. From personal observation while conducting field work in MWR, it is believed that the stocking rate of elephant

Stellenbosch University https://scholar.sun.ac.za

should fluctuate between 250 to 300 individuals. Therefore, the observed increase in the elephant population as well as the increase in other herbivore populations as revealed from the 2015 aerial survey suggests that MWR is overstocked and so the woody vegetation is under increasing pressure and is potentially being degraded.

Evidence of this was first reported by both Staub et al. (2013) and Weinand (2013) as both studies focused on the interaction of elephant with the woody vegetation in MWR. Staub et al. (2013) found that browsing intensity increased with proximity to water and that browsing was most concentrated in riparian woodland and least concentrated in Brachystegia-dominated woodland. Weinand (2013) found similar results and found that the most utilised water points were in the Sanctuary, where elephant browsing around the water points was high. Weinand (2013) found that the there was a positive but weak spatial relationship between elephant browsing and distance to water in this region due to the abundance of perennial water sources. The results from the current study noted that since 2013 the elephant population has increased and the water point usage by elephants is becoming saturated as there has been an increase in usage of the least preferred sites since 2013. Thus, indicating that due to the pressures of an increasing population density, the elephants have begun to utilise their least preferred areas, namely regions at high altitude with less preferred vegetation such as Brachystegia-dominated woodland, and so have increased their habitat usage to areas that previously did not incur a high elephant presence.

Evaluation of management options The non-equilibrium or flux of ecosystems is an integral part of ecosystem dynamics. External factors, such as fire, floods and droughts, are important in maintaining a diversity of plants and animals as well as generating space and a variety in ecosystems (Owen-Smith et al. 2006). This suggests that the disturbance caused by elephants could likewise be beneficial for ecosystem dynamics. However, it is not the local severity of elephant disturbance that is important but rather the spatial extent of it and so local overgrazing or over browsing by various large herbivore species is inconsequential or even beneficial if only a small portion of the land is affected (Gillson & Lindsay 2003, Owen-Smith et al. 2006). Concern arises when adverse effects occur and spread to cover large areas, decreasing the mosaic of the diversity of habitats (Whyte et al. 2003, Owen-Smith et al. 2006). The decrease in habitat heterogeneity is a reality for many fenced reserves with high elephant numbers and so management systems and plans are a necessity in order to preserve the remaining biodiversity in an area as well as maintain healthy populations of other herbivore species.

African Parks already have a number of effective management strategies in place in MWR and so only the relevant management suggestions for the elephant population will be discussed. It has been

Stellenbosch University https://scholar.sun.ac.za

established that the current elephant population in MWR is too high and needs to be controlled before it has any adverse effects on the vegetation and other herbivore species. However, it must be emphasised that elephant management requires a combination of prudent deliverable strategies that includes invasive management techniques, directly affecting the elephant population such as translocation, and non-invasive management techniques, indirectly affecting the population such as water point management. It is a combination of these practises that will obtain the optimum management results and so both invasive and non-invasive management practises will be discussed for MWR.

Translocation Translocation, an invasive management practise, reduces local elephant densities by removing elephants from high density to low density areas, establishing new populations or increasing existing populations in new conservation areas (Rogers & Sherwill 2008). The advantage of translocation is that it increases the range of elephants. The disadvantages are that it is a highly specialised and an extremely costly procedure that requires well-trained capture and transfer teams, as well as a variety of heavy duty equipment such as cranes and specialised transport (van Aarde & Jackson 2007; Rogers & Sherwill 2008). In addition, there are few conservation areas left in Southern Africa that can still accommodate extra elephants and even if individuals are translocated they still need to be managed, especially if the new area is fenced (van Aarde & Jackson 2007). From 1989 to 2001, a large scale translocation programme was completed in South Africa that translocated elephants from the Kruger National Park (hereafter, KNP) to 58 small fenced reserves. (Slotow et al. 2005). The average population growth rate in these small reserves averaged 8.3%, post translocation, indicating that the programme was largely successful (Slotow et al. 2005). This, however, increases the complexity of elephant management on these small reserves and many have either had to implement contraception programmes or continue with translocation in order to maintain population numbers at practical levels (Slotow et al. 2005).

Translocation is a primary management option for the elephant population in MWR. African Parks have planned to translocate 150 elephant from MWR to Nkhotakhota Wildlife Reserve, in central Malawi, in June/July 2017. Assuming that the elephant population consists of 389 individuals, according to the aerial survey, it will be decreased to 239 individuals post translocation, which is the recommended stocking rate for MWR. In order to promote the stabilisation of the population in terms of growth rate, it is recommended that the MWR population is maintained at 48% individuals above the age of 11 years and 6.8 % infants (<1year) (Calef 1988). Chapter 3, indicated that there are approximately 91 bulls (>11 years) and 38 elephant herds in MWR with an average herd size of 7 individuals, 2 adults (>11 years) and 5 young (<11 years). Using these averages, the suggested

Stellenbosch University https://scholar.sun.ac.za

proportion of individuals that should be translocated is estimated to be 70 young (<11 years) and 80 adults (>11 years) which varies between 14 -17 herds and a minimum of 30 bulls. These numbers will vary when actual herd sizes are taken into account. However, it is noted that if the sex ratio of the MWR elephant population is skewed towards bulls the population growth rate will decline (Whyte et al. 1998), thus it is recommended that fewer adult bulls than adult cows are translocated.

Contraception Contraception, an invasive management practise, manipulates the fertility in elephants using two main methods that have both been tested in Southern Africa, steroids and immunocontraception (Rogers & Sherwill 2008). Steroids have thus far been less successful than immunocontraception and so may be the less preferred method for MWR. Immunocontraception is a promising technique involving immunising cows with a foreign protein (porcine zona pellucida, pZP), and encouraging the production of antibodies that prevent fertilisation (Fayrer-Hosken et al. 2000; Rogers & Sherwill 2008). Field trials conducted in the Kruger National Park had success rates of 60% to 80% and have demonstrated that the method is effective in preventing pregnancies (Rogers & Sherwill 2008). The added advantage of pZP is that it is reversible, it has no adverse health effects, it is safe to use during pregnancy and has no behavioural side effects on elephant cows (Rogers & Sherwill 2008). Immunocontraception technology is continuously improving, the vaccine is delivered by darting and can be effective for a 12 month period but its long term effectiveness in controlling populations is still being evaluated (Fayrer-Hosken et al. 2000; Rogers & Sherwill 2008). There are many positives to elephant contraception including the possibility of lengthening inter-calving intervals and increasing the age of first calving (Pimm & van Aarde 2001). However, there are also limitations to contraception as it is incapable of reducing population sizes in the short term and to stabilise elephant numbers in large populations is very costly and time consuming (Whyte et al. 1998; Pimm & van Aarde 2001; van Aarde & Jackson 2007; Rogers & Sherwill 2008). To control the elephant population in the Kruger National Park by contraception would require the treatment of 2250 cows each year over an initial period of 11 years (Whyte et al. 1998). Therefore, contraception has both positive and negative attributes and on small fenced reserves may very well be the answer to increasing elephant populations as this management scheme would still be economically viable. However, on large reserves with large elephant populations, such as KNP, the cost of contraception may be too high as it an ongoing process that requires booster treatments (van Aarde et al. 1999; Fayrer-Hosken et al. 2000; Pimm & van Aarde 2001).

Once the translocation of 150 individuals in 2017 is complete, it is highly recommended that contraception is implemented in MWR. This is because the decrease in elephant density in MWR will encourage an increase in the population growth rate once again, and if contraception is not

Stellenbosch University https://scholar.sun.ac.za

implemented, it will be necessary to translocate another large number of individuals in the years to come (van Aarde & Jackson 2007; Chamaille-Jammes et al. 2008). It is estimated that the remaining number of adult females left in MWR will range from 50 -65 and if an annual population growth of 0% is required, immunocontraception models have demonstrated that contraception of 75% of these breeding-age females with an annual mortality rate of 2-3% will be sufficient (Delsink & Kirkpatrick 2012). It must be stressed that results expected, such as zero population growth rate, from the use of contraception may only be visible about 11 years from the initiation of the programme. This is due to the steady recruitment of females, born before the use of contraception, into the breeding population (Whyte 1998). The fertility rates, intercalving rates and mortality rates will affect the outcome of contraception and as these are unknown for the MWR population it is emphasized that contraception results are site specific and cannot be generalised (Whyte et al. 1998; van Aarde & Jackson 2007; Delsink & Kirkpatrick 2012).

Botanical Reserves and Zoning The establishment of botanical reserves, a non-invasive management technique, within fenced parks protects rare and endangered vegetation that is at high risk from elephant activities due to fencing (Rogers & Sherwill 2008). The KNP and the Addo Elephant National Park (hereafter, AENP) are both parks that have high elephant numbers that are having a detrimental effect on the biodiversity (Lombard et al. 2001; Whyte et al. 2003). Since the ban on culling in the KNP, management incorporated a new policy that divided the park into six management regions that included two botanical reserves where elephants were totally excluded, two high elephant impact zones and two low elephant impact zones where population numbers decrease at a rate of 7% per year (Whyte et al. 2003). AENP adopted a similar strategy with elephant exclusion zones to encourage botanical growth, these exclusion zones now contain plant species that are unable to grow elsewhere in the reserve due to high herbivory (Lombard et al. 2001). The limitations of this form of management are that it does not necessarily address the problem of high elephant numbers, the expense of installing and maintaining fences and it is mostly appropriate for small areas of highly threatened endemic vegetation (Rogers & Sherwill 2008). Botanical Reserves are a potential management solution for MWR in terms of vegetation preservation. However, there is a lack of information on the extent of vegetation degradation caused by elephant in MWR, as well as which plant species are being threatened. It is suggested that further research is conducted and thresholds of acceptable levels of elephant impacts on vegetation are determined before this management strategy is implemented.

Habitat Utilisation and Restriction: closing water points The positioning of water points, a non-invasive management technique, manipulates local elephant densities in large, water limited reserves (Chamaille-Jammes et al. 2007). The removal of any artificial

Stellenbosch University https://scholar.sun.ac.za

water sources located in areas away from rivers will effectively limit elephant foraging range in the dry season which has the potential to limit the impact on susceptible tree species that primarily grow at higher altitudes (Rogers & Sherwill 2008; Loarie et al. 2009). In addition, the overuse of areas close to water sources in the dry season, otherwise known as piosphere effects, will increase elephant’s foraging range and the greater distances moved within the dry season will supposedly increase calf mortality and decreasing population growth rates (Duffy et al. 2002; Owen-Smith et al. 2006; Rogers & Sherwill 2008). Chapter 3 suggested that elephant habitat use in the Sanctuary has become homogenised due to the overabundance of perennial water sources in the region. This supports the findings of Weinand (2013) who found that elephant browsing levels did not decrease with increasing distance from water in both seasons. The permanent closure of one artificial water hole in the Sanctuary is recommended, perhaps Nakamba or Nsepete, in order to reduce water availability in the region (Weinand 2013). The limitation is that MWR relies heavily on tourist revenue and the artificial waterholes in the Sanctuary are key tourist attractions as they provide great wildlife viewing opportunities in the dry season. The tourist experience may be negatively affected by crowding of people at remaining waterholes and so MWR will need to balance the needs of less mobile herbivore species and tourism with the management of elephant habitat use in an area.

Habitat utilisation and restriction is achieved through the combination of artificial waterhole locations and the implementation of zoning in the reserve which will require certain regions to be sacrificed for elephant use. The Sanctuary could potentially be designated as a high elephant impact zone due to the overabundance of perennial water sources as well as tourism demands. The remaining regions in the reserve would then be designated as low elephant impact zones and would contain fewer water sources at a minimum distance of 10km apart to reduce elephant habitat use (Weinand 2013). The concept of habitat heterogeneity with regards to elephant herbivory across MWR should be incorporated in the elephant management plan, promoting mosaics of different elephant browsing levels in different habitat types across MWR (Weinand 2013). This could be achieved by the strategic positioning of waterholes and the use of fencing to exclude elephants from certain areas (Owen-Smith et al. 2006; Chamaille-Jammes et al. 2007; O’Connor et al. 2007; Weinand 2013). As mentioned above, in order to implement zoning as a management tool in MWR further research needs to be conducted and decisions need to be made in terms of which areas will be zoned as high, low or no elephant impact (Weinand 2013).

Conclusion

In conclusion, the ecological effects of elephant can vary considerably between populations and depends on a number of factors that include rainfall, vegetation type and landscape features

Stellenbosch University https://scholar.sun.ac.za

(Guldemond & van Aarde 2008). In the present day, management can control the ecological effects or the intensity at which elephants utilise the landscapes through the use of invasive management techniques, such as contraception, and non-invasive management techniques, such as water provision. Spatial variability is most desirable in fenced reserves and in the modern day paradigm flux of conservation management it is preferred that elephant densities reflect a range of local densities which can be used as an index of potential habitat use intensity (Ferreira et al. 2012). Elephant populations respond to the variation in spatial and temporal resources in the short term through migration or dispersal and in the long term through changes in their demography (Young et al. 2009). In MWR it is recommended that management use both translocation and contraception in order to achieve the best possible results in controlling the population numbers. In addition, management should also employ the combination of zoning and water availability in order to promote spatial and temporal variability in elephant habitat use In MWR. However, elephants are long-lived mammals that have the longest life-history strategy of all terrestrial animals. This results in slow spatial and demographic population responses to management induced population regulations and restoration of ecological limitations and so ecological responses may take time to manifest.

As a parting word it is important to note that if reserve management only focus on reducing elephant numbers in protected areas it will only address the symptoms of an ecological problem that was initiated by management in the first place (van Aarde et al. 2007). In the majority of reserves fencing and artificial water supply are the root of the ‘elephant problem’ as they prevent the seasonal migration of elephants which, in unfenced systems, allows for vegetation recovery (van Aarde & Jackson 2007; van Aarde et al. 2007; Ferreira et al. 2012). Moreover, the provision of water, a key limiting factor in natural systems, reduces the natural mortality of elephants and so numbers become unnaturally high (van Aarde & Jackson 2007; van Aarde et al. 2007). Instead, reserve management need to focus on the issues that induce the symptoms of high elephant numbers in reserves, specifically fencing and water management, and the restoration of seasonal movement patterns of elephant (van Aarde & Jackson 2007). Therefore, the management of elephant should not solely focus on numbers but rather incorporate and focus on the idea of temporal and spatial scales.

Stellenbosch University https://scholar.sun.ac.za

References

 Barnes, R.F.W. (1983) The Elephant Problem in Rusha National Park, Tanzania. Biological Conservation, 26, 127-148.  Bothma, J., and J.G. Du Toit. (2010) Game ranch management. 5th edition. Van Schaik Publishers, Pretoria.  Calef, G.W. (1988) Maximum rate of increase in the African Elephant. African Journal of Ecology, 26, 323-327.  Caughley, G. (1976) The elephant problem – an alternative hypothesis. East African Wildlife Journal, 14, 265-283.  Chamaille-Jammes, S., Fritz, H., Valeix, M., Murindagomo, F. & Clobert, J. (2008) Resource variability, aggregation and direct density dependence in an open context: the local regulation of an African elephant population. Journal of Animal Ecology, 77, 135-144.  Chamaille-Jammes, S., Valeix, M. & Fritz, H. (2007) Managing heterogeneity in elephant distribution: interactions between elephant population density and surface water availability. Journal of Applied Ecology, 44, 625-633.  Delsink, A.D. & Kirkpatrik, J.F. (2012) Free-ranging African Elephant Immunocontraception a new paradigm for elephant management. Trident Press, Cape Town, South Africa.  Duffy, K.j., van Os, R., Vos, S., van Aarde, J. Ellish, G. & Stretch, A-M.B. (2002) Estimating impact of reintroduced elephant on trees in a small reserve. South African Journal of Wildlife Research, 32(1), 23-29.  Fayrer-Hosken, R.A., Grobler, D., Van Altena, J.J., Bertschinger, H.J. & Kirkpatrick, J.F. (2000) Immunocontraception of African elephants: A humane method to control elephant populations without behavioural side effects. Nature, 407, 149-150.  Ferreira, S.M., Freitag-Ronalson, S., Pienaar, D. & Hendriks, H. (2012) Elephant Management Plan: Kruger National Park 2013-2022. Scientific Services, SANParks, Skukuza. South Africa.  Gillson, L. & Lindsay, K. (2003) Ivory and ecology – changing perspectives on elephant management and the international trade in ivory. Environmental Science, 6, 411-419.  Guldemond, R. & Van Aarded, R. (2008) A Meta-Analysis of the Impact of African Elephants on Savanna Vegetation. Journal of Wildlife Management, 72(4), 892-899.  Kalwij, J.M., de Boer, W.F., Mucina, L., Prins, H.H.T., Skarpe, C. & Winterbach, C. (2010) Tree cover and biomass increase in a Southern African savanna despite growing elephant population. Ecological Application, 20(1), 222-233.

Stellenbosch University https://scholar.sun.ac.za

 Kerley, G.I.H. & Landman, M. (2006) The impacts of elephants on biodiversity in the Eastern Cape Subtropical Thickets. South African Journal of Science, 102, 395-402.  Loarie, S.R., van Aarde, R.J. & Pimm, S.L. (2009) Fences and artificial water affect African savannah elephant movement patterns. Biological Conservation, 142(12), 3086-3098.  Lombard, A.T., Johnson, C.F., Cowling, R.M. & Pressey, R.L. (2001) Protecting plants from elephants: botanical reserve scenarios within the Addo Elephant National Park, South Africa. Biological Conservation, 102, 191-203.  O’Connor, T., Goodman, P.S. & Clegg, B. (2007) A functional hypothesis of the threat of local extirpation of woody plant species by elephant in Africa. Biological Conservation, 136, 329- 345.  Owen-Smith, N., Kerley, G.I.H., Page, B., Slotow, R. & van Aarde, R.J. (2006) A scientific perspective on the management of elephants in the Kruger National Park and elsewhere. South African Journal of Science, 102, 389-394.  Pimm, S.L. & van Aarde, R.J. (2001) African elephants and contraception. Nature, 411, 766- 767.  Rogers, K. & Sherwill, T. (2008) Towards a new elephant management policy for South Africa (1995-2008). Conservation Services Division. Pretoria, South Africa.  Shoshani, J., Hagos, Y., Yacob, Y., Ghebrehiwet, M. & Kebrom, E. (2004) Elephants (Loxodonta africana) of Zoba Gash Barka, Eritrea: Part 2. Numbers and distribution, ecology and behaviour and, fauna and flora in their ecosystem. Pachyderm, 36, 52-68.  Shrader, A.M., Pimm, S.L, & van Aarde, R.J. (2010) Elephant survival, rainfall and the confounding effects of water provision and fences. Biodiversity Conservation, 19, 2235-2245.  Skarpe, C., Aarrestad, P.A., Andreassen, H.P., Dhillion, S.S., Dimakatso, T., du Toit, J.T., Halley, D.J., Hytteborn, H., Makhabu, S., Mari, M., Marokane, W., Masunga, G., Modise, D., Moe, S.R., Mojaphoko, R., Mosugelo, D., Motsumi, M., Neo-Mahupeleng, G., Ramotadima, M., Rutina, L., Sechele, L. & Sejoe, T.B., Stokke, S., Swenson, J.E., Taolo, C., Vandewalle, M. & Wegge, P. (2004) The return of the giants: Ecological Effects of an Increasing Elephant Population. Ambio, 33(6), 276-282.  Slotow, R., Garai, M.E., Reilly, B., Page, B.& Carr, R.D. (2005) Population dynamics of elephants re-introduced to small fenced reserves in South Africa. South African Journal of Wildlife Research, 35(1), 23-32.  Staub, C.G., Binford, M.W. & Stevens F.R. (2013) Elephant herbivory in Majete Wildlife Reserve, Malawi. African Journal of Ecology, 51(4), 536-543.

Stellenbosch University https://scholar.sun.ac.za

 Styles, C.V. & Skinner, J.D. (2000) The influence of large mammalian herbivores on growth form and utilization of mopane trees, Colophospermum mopane, in Botswana’s Northern Tuli Game Reserve. African Journal of Ecology, 38, 95-101.  van Aarde, R.J. & Jackson, T.P. (2007) Megaparks for metapopulations: Addressing the causes of locally high elephant numbers in Southern Africa. Biological Conservation, 134, 289-297.  van Aarde, R.J., Whyte, I. & Pimm, S. (1999) Culling and the dynamics of the Kruger National Park African elephant population. Animal Conservation, 2, 287-294.  Van Aarde, R.J., Ferreira, S. & Wassenaar, T. (2007) Elephant Management: There is more to it than meets the eye. Conservation, 1-10.  van Oudtshoorn, F. (2015) Veld Management, Principles and Practices. 1st edn. Briza Publications. Pretoria, South Africa.  Weinand, J.J. (2013) Woody vegetation change and elephant water point use in Majete Wildlife Reserve: implications for water management strategies. MSc thesis, University of Stellenbosch, Stellenbosch, South Africa.  Whyte, I., van Aarde, R. & Pimm, S.L. (1998) Managing the elephants of Kruger National Park. Animal Conservation, 1, 77-83.  Whyte, I.J., van Aarde, R.J. and Pimm, S.L. (2003) Kruger’s Elephant Population: Its size and Consequences for Ecosystem Heterogeneity. The Kruger Experience. (eds J.T. du Toit, K.H. Rogers & H.C. Biggs), pp: 332-348. Island Press. Washington.  Young, K.D., Ferreira, S.M. & van Aarde, R.J. (2009) Elephant spatial use in wet and dry savannas of Southern Africa. Journal of Zoology, 278(3), 189-205.

100

Stellenbosch University https://scholar.sun.ac.za

Appendices

Appendix 1

The wildlife reintroduction history of Majete Wildlife Reserve

Wildlife was reintroduced into Majete Wildlife Reserve in stages, pre 2008 all introductions were made into the Sanctuary as it was the only region that had a complete fence. Total fencing was completed in 2008 and further wildlife reintroductions were made to the rest of the reserve. The Sanctuary fence was only removed in 2011.

Table A.1. Details of wildlife reintroductions to various regions of Majete Wildlife Reserve between 2003 and 2010. Sanctuary The rest of Majete Species 2003 2004 2005 2006 2007 2008 2009 2008 2009 2010 Total Elephant 70 60 83 213 Black Rhino 2 6 8 Buffalo 120 100 86 306 Sable 100 153 99 352 Hartebeest 4 10 15 30 59 Waterbuck 98 198 106 402 Zebra 37 50 9 38 40 174 Eland 20 32 25 77 Impala 216 210 311 737 Nyala 6 15 38 59 Warthog 60 98 158

101

Stellenbosch University https://scholar.sun.ac.za

Appendix 2

Appendix 2.1

An example of the Majete Elephant Database. The database was designed with a pyramid of folders, the first layer focuses on the gender of an individual or different regions herd frequent. The second layer focuses on the anatomy of individuals, specifically looking at notable characteristics of each one. The final layer lists the categories of characteristics, for example, the shapes of scars on the ears were listed in the figure below. By identifying notable characteristics of individuals met in the field one can later identify them in the database. The same applies to herds.

102

Stellenbosch University https://scholar.sun.ac.za

Appendix 2.2

An example of a well-known family herd in MWR. Each of the adult cows was given a unique code such as ‘C8’ which stands for ‘Cow 8’. When each herd was recorded in the database, the name given to the herd was composed of the adult females’ codes as well as the number of individuals in a herd. The finer data recorded for each herd specifically looked at all the individuals in a herd, aging them and sexing them.

Adult Cow (C8) - Toothless Adult Cow (C9) Adult Cow (C10)

Juvenile Female (C26) Calf Male Calf Male

Calf Female Infant Male Infant Female

103

Stellenbosch University https://scholar.sun.ac.za

Appendix 2.3

A.2. A Table representing the number of family herds and solitary bulls in MWR and their total number, sex and age structure in the years 2015 and 2016.

Code Group Type Total Number of Number of Number of Number of Number of Number of Number of Number of Number of Number of Number of Number of (Code allocated to (Family Number Large Medium Small Number Male Female Unknown Male Female Unknown Male Female Unknown each individual eg: Herd, of Adult Adult Adult Juveniles Juveniles Juveniles Calves Calves Calves Infants Infants Infants C1 = Cow One) Solitary Bull) Individuals Females Females Females

1 C1_C2_C13 Family Herd 10 0 3 0 0 1 0 3 0 0 1 2 0

2 C3_C7_C14 Family Herd 7 0 1 1 0 1 0 1 2 0 1 0 0 C4_C5_C6_C15_ 5 Family Herd 15 1 2 2 0 2 0 3 1 2 1 0 1 C17 6 C8_C9_C10_C26 Family Herd 10 0 2 2 1 0 0 1 2 0 1 1 0

7 C11_C12 Family Herd 7 0 1 1 0 0 0 1 2 0 0 0 2

8 C19 Family Herd 6 0 2 0 0 0 0 0 0 2 0 0 2

9 C20_C21_C22_C23 Family Herd 11 1 2 2 1 1 2 0 1 1 1 0 0

10 C24_C25_C29 Family Herd 8 3 0 0 1 1 0 2 0 0 0 0 1

13 C32_C33 Family Herd 7 1 1 0 0 1 1 0 0 1 0 0 2

14 C34_C35_C36 Family Herd 10 2 1 1 1 1 0 1 0 2 0 0 1

15 C37_C38_C39_C40 Family Herd 12 0 3 1 1 2 0 1 1 2 0 0 1

18 PwC1_PwC29 Family Herd 4 0 1 1 0 0 0 1 0 1 0 0 0

19 PwC2 Family Herd 5 1 0 0 0 1 0 1 1 0 0 0 1

20 PwC3 Family Herd 6 0 1 1 1 0 0 0 0 2 0 0 1 PwC5_PwC6_ 21 Family Herd 11 0 2 1 2 1 0 2 1 0 0 1 1 PwC15 22 PwC7_PwC8 Family Herd 7 1 1 1 1 0 0 1 1 0 0 0 1

104

Stellenbosch University https://scholar.sun.ac.za

23 PwC9 Family Herd 4 0 1 0 1 0 0 1 0 0 0 0 1

24 PwC10 Family Herd 5 1 0 0 1 0 0 0 0 2 0 0 1

25 PwC11_PwC14 Family Herd 7 0 2 0 0 1 0 0 0 2 0 0 2

26 PwC12_PwC13 Family Herd 8 1 1 0 1 1 0 0 0 3 0 0 1

27 PwC16 Family Herd 4 0 1 0 0 0 0 1 0 1 0 0 1

28 PwC17 Family Herd 5 0 1 0 1 0 0 0 0 2 0 0 1

29 PwC18 Family Herd 5 0 1 0 1 1 0 0 0 1 0 0 1

30 PwC19_PwC22 Family Herd 6 0 1 1 1 0 0 0 1 1 0 0 1

31 PwC20 Family Herd 3 1 0 0 1 0 0 0 0 1 0 0 0

32 PwC23 Family Herd 4 0 1 0 0 1 0 1 0 0 0 1 0

33 PwC24 Family Herd 2 0 1 0 0 0 0 1 0 0 0 0 0 PwC25_PwC26_ 34 Family Herd 8 0 1 2 0 2 0 0 0 2 0 0 1 PwC27 36 PeC1_PeC2 Family Herd 9 0 2 1 0 0 2 1 1 1 0 0 1

38 PeC5_PeC17 Family Herd 7 1 1 0 1 0 0 0 2 1 0 0 0 PeC6_PeC7_PeC8_ 39 Family Herd 11 1 2 1 0 1 1 2 1 0 0 0 2 PeC14 40 PeC9_PeC16 Family Herd 6 0 2 0 0 1 0 0 0 2 0 0 1 PeC10_Pe 41 C11_PeC12_PeC13_ Family Herd 12 1 2 2 1 1 0 1 1 2 0 0 1 PeC15 PeC19 _ PeC22_ 42 Family Herd 8 1 2 0 1 1 0 0 0 2 0 0 1 PeC27 43 PeC20_PeC21 Family Herd 9 1 1 0 0 0 2 0 0 2 0 0 2

44 PeC23_PeC24 Family Herd 4 0 2 0 0 0 1 0 0 1 0 0 0

46 DiC2 Family Herd 8 2 1 0 1 0 0 0 1 2 0 0 1

105

Stellenbosch University https://scholar.sun.ac.za

47 DiC4_DiC5 Family Herd 5 0 1 1 1 1 0 0 0 1 0 0 0

48 B1 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

49 B2 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

50 B3 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

51 B4 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

52 B5 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

53 B6 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

54 B7 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

55 B8 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

56 B9 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

57 B10 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

58 B11 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

59 B12 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

60 B13 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

61 B14 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

62 B15 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

63 B16 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

64 B17 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

65 B18 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

66 B19 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

67 B20 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

106

Stellenbosch University https://scholar.sun.ac.za

68 B21 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

69 B22 Bull 1 1 0 0 0 0 0 0 0 0 0 0 0

70 B23 Bull 1 1 0 0 0 0 0 0 0 0 0 0 0

71 B24 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

72 B25 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

73 B26 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

74 B26 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

75 B27 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

76 B28 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

77 B29 Bull 1 1 0 0 0 0 0 0 0 0 0 0 0

78 B30 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

79 B31 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

80 B32 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

81 B33 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

82 B34 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

83 B35 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

84 B36 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

85 B37 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

86 B38 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

87 B39 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

88 B40 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

107

Stellenbosch University https://scholar.sun.ac.za

89 B41 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

90 B42 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

91 B43 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

92 B44 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

93 B45 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

94 B46 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

95 B47 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

96 B48 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

97 B49 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

98 B50 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

99 B51 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

100 B52 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

101 PwB1 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

102 PwB2 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

103 PwB3 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

104 PwB4 Bull 1 1 0 0 0 0 0 0 0 0 0 0 0

105 PwB5 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

106 PwB6 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

107 PwB7 Bull 1 1 0 0 0 0 0 0 0 0 0 0 0

108 PwB8 Bull 1 1 0 0 0 0 0 0 0 0 0 0 0

109 PwB9 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

108

Stellenbosch University https://scholar.sun.ac.za

110 PwB10 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

111 PwB11 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

112 PwB12 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

113 PwB13 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

114 PwB14 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

115 PwB15 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

116 PwB16 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

117 PwB17 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

118 PwB18 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

119 PwB19 Bull 1 1 0 0 0 0 0 0 0 0 0 0 0

120 PwB20 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

121 PwB21 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

122 PeB1 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

123 PeB2 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

124 Pe B3 Bull 1 1 0 0 0 0 0 0 0 0 0 0 0

125 Pe B4 Bull 1 1 0 0 0 0 0 0 0 0 0 0 0

126 Pe B5 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

127 Pe B6 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

128 Pe B7 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

129 Pe B8 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

130 Pe B9 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

109

Stellenbosch University https://scholar.sun.ac.za

131 Pe B10 Bull 1 1 0 0 0 0 0 0 0 0 0 0 0

132 Pe B11 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

133 Pe B13 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

134 Pe B14 Bull 1 0 0 1 0 0 0 0 0 0 0 0 0

135 Pe B15 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

136 Pe B16 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

137 Pe B17 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

138 Pe B18 Bull 1 0 1 0 0 0 0 0 0 0 0 0 0

Total 366 30 98 55 21 23 9 26 19 42 5 5 33

110